<<

Rocky Mountain National Park, 2001-2005 VEGETATION CLASSIFICATION AND MAPPING

FINAL REPORT --August, 2005

Technical Memorandum 8260-05-02 Remote Sensing and GIS Group Technical Service Center Bureau of Reclamation , Colorado This report was prepared for the U.S. National Park Service and the U.S. Geological Survey’s Center for Biological Informatics by the Remote Sensing and Geographic Information Group of the Bureau of Reclamation’s Technical Service Center, Denver, Colorado in cooperation with Colorado Natural Heritage Program and NatureServe.

Technical Memorandum No. 8260-05-0X.

U.S. Department of the Interior Mission Statement

The mission of the Department of the Interior is to protect and provide access to our Nation’s natural and cultural heritage and honor our trust responsibilities to tribes.

Mission of U. S. Bureau of Reclamation

The mission of the Bureau of Reclamation is to manage, develop, and protect water and related resources in an environmentally and economically sound manner in the interest of the American public.

The Remote Sensing and Geographic Information Group, organized in 1975, provides assistance and advice regarding the application of remote sensing and geographic information systems (GIS) technologies to meet the spatial information needs of the Bureau of Reclamation and other governmental clients.

USGS-NPS VEGETATION MAPPING PROGRAM Rocky Mountain National Park, Colorado

David Salas Physical Scientist Bureau of Reclamation Remote Sensing and GIS Group Denver, Colorado

Joe Stevens Ecology Team Leader Colorado Natural Heritage Program Fort Collins, Colorado

Keith Schulz Regional Vegetation Ecologist NatureServe Boulder, CO

Report Produced by: Program Managed by: U.S. Bureau of Reclamation U.S. Geological Survey Technical Service Center Center for Biological Informatics Remote Sensing and GIS Group Denver Federal Center, Building 810 Mail Code D-8260 Denver Federal Room 8000, MS 302 Center Building 56 Denver, Colorado 80225-0046 Denver, Colorado 80225

Cooperation agencies and organizations:

iii USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

TABLE OF CONTENTS

LIST OF TABLES...... VII

LIST OF FIGURES...... VIII

LIST OF CONTACTS AND CONTRIBUTORS ...... IX

ACKNOWLEDGEMENTS ...... XIII

LIST OF ABBREVIATIONS AND ACRONYMS...... XV

EXECUTIVE SUMMARY ...... 1

INTRODUCTION ...... 3

BACKGROUND ...... 3 USGS-NPS Park Vegetation Mapping Program ...... 3 ROCKY MOUNTAIN NATIONAL PARK VEGETATION MAPPING PROJECT ...... 4 PREVIOUS VEGETATION MAPS ...... 5 SCOPE OF WORK...... 6 INTRODUCTION TO THE NATIONAL VEGETATION CLASSIFICATION (NVC) AND SYSTEM (NVCS)...... 8 INTRODUCTION TO NATURAL HERITAGE PROGRAM METHODOLOGY AND ELEMENT RANKING ...... 11 The Natural Heritage Ranking System ...... 11 Legal Designations for Rare Species...... 13 Element Occurrences and their Ranking...... 13

PROJECT AREA...... 15

LOCATION AND REGIONAL SETTING...... 15 CLIMATE AND WEATHER...... 18 TOPOGRAPHY ...... 18 GEOLOGY ...... 20 Precambrian ...... 20 Paleozoic...... 21 Mesozoic...... 21 Cenozoic ...... 21 Quaternary ...... 21 SOILS ...... 22 WILDLIFE ...... 23 HYDROLOGY...... 23 VEGETATION ...... 25 Overview...... 25 East/West Slope Effects ...... 25 Alpine Zone...... 27 Subalpine Zone...... 28 Montane Zones ...... 29 Wetland and Riparian...... 29

METHODS...... 31

PLANNING AND SCOPING ...... 31 BOR Responsibilities and Deliverables: ...... 31 NatureServe Responsibilities and Deliverables: ...... 31 CNHP Responsibilities and Deliverables:...... 32 Scoping Meetings ...... 32 Initial Planning Meetings: ...... 32

iv USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Gradsect Meeting:...... 32 Field Preparation Meetings: ...... 33 Interim Status Meeting:...... 33 Map Unit Meeting:...... 33 Accuracy Assessment Meeting: ...... 33 PRELIMINARY DATA COLLECTION AND REVIEW OF EXISTING INFORMATION...... 33 FIELD SURVEY...... 36 Sampling Design: Stratified Random Gradsect...... 36 General Plot Collection Considerations ...... 37 Data Collection: Relevé Plots ...... 38 Data Collection: Forest Fuels Data...... 39 AERIAL PHOTOGRAPHY ACQUISITION AND ORTHOBASE MAP DEVELOPMENT ...... 41 PHOTO INTERPRETATION – MAP UNITS...... 44 Photo Interpretation ...... 44 Map Units and Polygon Attribution ...... 44 DIGITAL TRANSFER ...... 48 PLOT DATA MANAGEMENT AND CLASSIFICATION ANALYSIS ...... 49 Plot Data Management...... 49 Vegetation Classification...... 49 MAP VERIFICATION AND ACCURACY ASSESSMENT...... 51 Sample Design...... 51 Sample Method ...... 51 Sample Site Selection...... 51 Data Collection – AA Points ...... 52 AA Metrics...... 57 Binary accuracy assessment:...... 57 Confidence Interval:...... 57 Fuzzy Accuracy Assessment:...... 58 Hypothesis Testing...... 61 Error Distribution...... 62

RESULTS ...... 64

FIELD DATA COLLECTION ...... 64 VEGETATION CLASSIFICATION ...... 64 PHOTOGRAPHIC DATABASE ...... 65 PHOTO-INTERPRETATION AND MAP UNITS ...... 93 VEGETATION MAP ...... 102 FUZZY ACCURACY ASSESSMENT...... 106 Accuracy Assessment by Map Unit...... 113 Fuzzy Error Distribution ...... 122

DISCUSSION ...... 124

NVC CLASSIFICATION...... 124 Global Rarity...... 125 Non-native Species ...... 127 AERIAL PHOTOGRAPHY AND ORTHOPHOTOS...... 128 PHOTO-INTERPRETATION AND MAP UNITS ...... 128 MAP ACCURACY...... 129 General Considerations...... 129 The Accuracy Standard for the NBS/NPS Vegetation Mapping Program...... 131 RECOMMENDATIONS ...... 133 Field Survey...... 133 Vegetation Map ...... 134 Map Units...... 134 Accuracy...... 134 Map Improvement Suggestions:...... 135

v USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

BIBLIOGRAPHY...... 137

PROJECT DVD-ROM ...... 146

vi USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

LIST OF TABLES TABLE 1. SUMMARY STATISTICAL COMPARISON OF CURRENT MAP EFFORT TO EXISTING VEGETATION MAP...... 5 TABLE 2. SUMMARY OF THE NATIONAL VEGETATION CLASSIFICATION SYSTEM HIERARCHAL APPROACH (MAYBURY 1999) ...... 8 TABLE 3. DEFINITION OF NATURAL HERITAGE IMPERILMENT RANKS ...... 12 TABLE 4. ELEMENT OCCURRENCE RANKS AND THEIR DEFINITIONS...... 14 TABLE 5. LIFE ZONES AS DESCRIBED BY MERRIAM AND STEINEGER (1890) AND THEIR MODERN EQUIVALENT ...... 26 TABLE 6. ENVIRONMENTAL VARIABLES AND CLASSES USED IN THE MODIFIED GRADSECT ANALYSIS FOR ROMO...... 37 TABLE 7. STRUCTURAL CATEGORIES FOR VEGETATION PHOTOINTERPRETATION...... 46 TABLE 8. POLYGON ATTRIBUTE ITEMS AND DESCRIPTIONS USED IN THE ROMO SPATIAL DATABASE (GIS COVERAGE) ...... 47 TABLE 9. RECOMMENDED MAP ACCURACY SAMPLE NUMBER PER CLASS BY FREQUENCY AND AREA ...... 55 TABLE 10. FUZZY SET ACCURACY RANKS (GOPAL AND WOODCOCK, 1994)...... 60 TABLE 11. VEGETATION CLASSIFICATION FOR ROMO ...... 67 TABLE 12. MAP UNITS AND VEGETATION ASSOCIATIONS WITHIN EACH MAP UNIT ...... 94 TABLE 13. SPATIAL STATISTICS FOR ALL MAP UNITS IN THE MAPPING AREA ...... 103 TABLE 14. SUMMARY AREA STATISTICS FOR MAP UNITS WITHIN ROMO ...... 105 TABLE 15. OVERALL MAP ACCURACIES FOR EACH FUZZY CLASS ...... 107 TABLE 16. CONTINGENCY TABLE FOR FUZZY ACCURACY ASSESSMENT LEVEL 5...... 108 TABLE 17. CONTINGENCY TABLE FOR FUZZY ACCURACY ASSESSMENT LEVEL 4...... 109 TABLE 18. CONTINGENCY TABLE FOR FUZZY ACCURACY ASSESSMENT LEVEL 3...... 110 TABLE 19. MAP UNIT ACCURACIES FOR OMISSION AND COMMISSION ERRORS FOR ALL ERRORS AT ALL FUZZY LEVELS (ONLY VEGETATED POLYGONS) ...... 111 TABLE 20. CHANGE IN MEAN MAP UNIT ACCURACIES BY FUZZY LEVEL...... 112 TABLE 21. NATURAL HERITAGE ELEMENTS IDENTIFIED FROM PLOT DATA ...... 126 TABLE 22. COMPARATIVE FUZZY ACCURACIES FOR FOUR NATIONAL PARKS...... 132

Terra Tomah

vii USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

LIST OF FIGURES

FIGURE 1. VEGETATION MAPPING – PROJECT AND MAP BOUNDARIES ...... 7 FIGURE 2. LOCATION MAP – ROMO ...... 16 FIGURE 3. DETAIL MAP – ROMO ...... 17 FIGURE 4. TOPOGRAPHY OF ROCKY NATIONAL PARK END ENVIRONS...... 20 FIGURE 5. LOCATION OF ANCILLARY PLOTS USED FOR PHOTOINTERPRETATIVE TRAINING AND REFERENCE ...... 35 FIGURE 7. FLIGHT LINES AND PHOTO CENTERS FOR 1:12,000 COLOR PHOTOGRAPHY ...... 42 FIGURE 8. FLIGHT LINES AND PHOTO CENTERS FOR 1:40,000 COLOR PHOTOGRAPHY ...... 43 FIGURE 9. EXAMPLE OF TRANSFER PROCESS FROM AERIAL 9 X 9 PHOTOGRAPHY TO COLOR ORTHOPHOTO BASEMAP 46 FIGURE 10. LOCATION OF VEGETATION PLOTS COLLECTED AT ROMO...... 54 FIGURE 11. DISTRIBUTION OF ACCURACY ASSESSMENT POINTS, ROMO AND ENVIRONS...... 56 FIGURE 12. EXAMPLE OF FUZZY RE-DESIGNATION FOR THE CREATION OF CONTINGENCY TABLES ...... 61 FIGURE 13. SEMIVARIOGRAM AND SPHERICAL MODEL FOR 1207 AA POINTS IN ROMO...... 63 FIGURE 14. FUZZY SPATIAL VIEW OF ACCURACY ...... 123

Big Horn Sheep

viii USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

LIST OF CONTACTS AND CONTRIBUTORS

U. S. Department of the Interior Geological Survey - Biological Resources Division

Karl Brown Program Coordinator - USGS-NPS Vegetation Mapping Program U.S. Geological Survey Center for Biological Informatics P.O. Box 25046 Denver, Colorado 80225-0046 Phone (303) 202-4240 E-Mail: [email protected] Website: http://biology.usgs.gov/cbi

Krumholz

Elk

ix USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

U. S. Department of the Interior U.S. National Park Service

Mike Story Jeff Conner NPS Program Coordinator Research and Resource Management Supervisor NPS-WASO Natural Resource Information Natural Resources Specialist, Vegetation Division I&M Program Mapping Coordinator, Rocky Mountain National Phone: 303-969-2746 Park, Estes Park, CO. 80517, (970) 586-1296 E-mail: [email protected] Website: http://www.nps.gov/romo Website: http://biology.usgs.gov/npsveg/index.html Ron Thomas GIS Specialist, Vegetation Mapping Chris Lea Coordinator, Rocky Mountain National Park, NPS Botanist / Ecologist Estes Park, CO. 80517, (970) 586-1292 P.O. Box 25287 [email protected] Denver, CO 80225-0287 Phone: (303) 969-2807 Nathan Williamson

E-mail: [email protected] Fire Effects Specialist, Vegetation Mapping Coordinator, Rocky Mountain National Park, Estes Park, CO, 80517 (970) 586-1434

x USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

U. S. Department of the Interior

Bureau of Reclamation

Michael Pucherelli Trudy Meyer Group Manager Lead GIS Specialist Remote Sensing and Geographic Phone: 303-445-2278 Information Group Email: [email protected] Mail Code D-8260 Denver Federal Center Building 56 Dan Cogan Denver, Colorado 80225 Contractor Phone: (303) 445-2267 Phone: (303) 910-9768

E-mail: [email protected] E-mail: [email protected]

Rock spires along the Continental Divide, near , in Rocky Mountain National Park. Vertical fractures have been enhanced by frost action and weathering during Quaternary ice ages. Horizontal banding shows light-colored granite intruded into metamorphic rocks. This is a popular rock-climbing area. Photo by Jim Cole in 1971 (From http://www.aese.org/a_annual_meetingFieldTrip.html).

xi USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Precipitous cliff walls of the east face of , as viewed from Chasm Lake, some 2,400 feet below the summit of the mountain. At 14,255 feet, Longs Peak is the highest peak in Rocky Mountain National Park. It is a prominent landmark visible from Denver and the surrounding . Photograph by W.T. Lee on July 22, 1916. U.S. Geological Survey Photographic Library (From http://www.aese.org/a_annual_meetingFieldTrip. html). This view is similar to the cover photograph but taken above Peacock Pool.

Sprague Lake

xii USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

ACKNOWLEDGEMENTS

A project of this size, complexity, expanse Michael Schindel of The Nature and onerous logistics necessarily required Conservancy provided the gradsect the enthusiasm and energy of many people analysis; over several years. The dedication of all involved helped to produce a product that Dan Cogan, Sabrina Chartier and Dave we, the authors, gratefully acknowledge. McMullen helped with the The combination of ecologists, photointerpretation; geographers, botanists and natural resource professionals in all the Karl Brown of the USGS and Mike Story cooperating agencies and organizations with the NPS for bringing this project to allowed for a phenomenal work the BOR and then being there for help environment that greatly reduced the with coordination, logistics, and financial stresses one might imagine for a project of matters; this size. Many thanks go to Trudy Meyer and Kurt We would specifically like to thank the Wille of the BOR GIS staff. Trudy for following people for their efforts: making sense of all the line-work and hieroglyphic labels to produce a very large Gwen Kittel from NatureServe for her digital map and Kurt for developing in- help training the 2002 and 2004 field house programs to ease several steps of the crews on the nuances of willows. In entire process; addition, Jim Drake, Don Faber- Langendoen, and Marion Reid from The 2004 Accuracy Assessment field crew NatureServe helped lead a week-long for getting more plots than we needed, regional NPS Vegetation Mapping despite the wetter and colder weather than Program field methods training in May expected: Kelsey Forrest, Katie Driver, 2002 for field crews from several parks Rhiannon Thomas, Karin Edwards, Brett including the ROMO field crews. Wolk, and Shannon Gilpin;

Nate Williamson was instrumental for his Extra special attention and thanks goes to input to the field data collection both Jeff Connor and Ron Thomas of requirements and methods for Fire Pro. Rocky Mountain National Park. Jeff’s steady hand and Ron’s cryptic humor were The 2002 vegetation field crew for their really instrumental in the success of this dedication to collecting good quality plot project. The GIS/GPS support, maps, and data despite the sometimes cold, wet photo organization Ron provided were weather and all the long difficult days in critical to the ability of the crews to get in the field: Joanna Lemly, Melissa the field and get their work done. Without Hornbein, Donna Shorrock, Justin West, his help we would still be out there trying Sabrina Chartier, Jessica Miesel, Lynn to find our way. Wheeler, and Jennifer Jones; We thank Rhianon Thomas, of the accuracy assessment field crew, for the

xiii USGS-NPS Vegetation Mapping Program Rocky Mountain National Park cover photograph and some text, the public Billy Schweiger and Daniel Manier of the domain “Colorado Headwater” photograph National Park Service, Rocky Mountain by Jeff Marmorstone, and the “Rock spires I&M Network. Their time and comments along the Continental Divide” photograph on both the report and key are greatly by Jim Cole. All other photographs in the appreciated. main report courtesy of Rocky Mountain National Park website. Photographs in the We would also be remiss if we did not also appendices are from various members of acknowledge the efforts of both Mike Story the field crews. and Karl Brown that have been sheparding these park vegetation mapping projects for Alan Bell of the RSGIS Denver provided so long now. Along with Mike and Karl the very rigorous and time consuming are the many other scientists that produced review of all the photographic and digital the original protocols and those that have products contracted to Horizons. been in the field applying these techniques to projects for many many years. These Katie Neuhaus of CNHP for her attention previous efforts have raised the bar each to detail and her many and varied skills that time. We have learned much from the made the final product better than it would earlier projects and it our distinct pleasure have been otherwise. to have added add to them. New technologies and lessons learned have, we Janet Coles was a big help during the hope, produced a finer product. gradsect and map unit discussions. A final field test of the dichotomous key to the David, Joe and Keith vegetation associations was done by

xiv USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

LIST OF ABBREVIATIONS AND ACRONYMS

AA Accuracy Assessment AML Arc Macro Language BOR Bureau of Reclamation (also USBR) BRD Biological Resource Division (of the USGS) CBI Center for Biological Informatics (of the USGS/BRD) CIR Color Infrared Photography CNHP Colorado Natural Heritage Program DEM Digital Elevation Model DLG Digital Line Graph DRG Digital Raster Graphic DOQQ Digital Orthophoto Quarter Quadrangle FGDC Federal Geographic Data Committee GIS Geographic Information System(s) GPS Global Positioning System IMU Inertial Measurement Unit MMU Minimum Mapping Unit NPS U.S. National Park Service NAD North American Datum NBII National Biological Information Infrastructure NRCS Natural Resources Conservation Service (formerly the Soil Conservation Service - SCS) NREL Natural Resource Ecology Lab, Colorado State University, Fort Collins. NVC National Vegetation Classification NVCS National Vegetation Classification System NRCS Natural Resource Conservation Service (Formerly – SCS) NWI National Wetland Inventory PARK ROMO RMS Root Mean square RSGIG Remote Sensing and Geographic Information Group TNC The Nature Conservancy USBR United States Bureau Of Reclamation (also BOR) USDA-SCS U.S. Dept. Of Agriculture – Soil Conservation Service USFS United States Forest Service USGS United States Geological Survey UTM Universal Transverse Mercator ROMO Rocky Mountain National Park

xv USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

EXECUTIVE SUMMARY

Rocky Mountain National Park (ROMO) landuse at ROMO. In the end, 46 map encompasses 417 square miles in north units were developed and directly cross- , and lies on the eastern walked or matched to corresponding slope of the southern . plant associations and land-use classes. This mapping effort is part of the All of the interpreted and remotely National Park Services’ national sensed data were converted to inventory and monitoring program and Geographic Information System (GIS) will provide core or ‘baseline’ databases using ArcInfo© software. information that park managers need to Draft maps created from the vegetation effectively manage and protect park classification were field-tested and resources. This vegetation inventory was revised before independent ecologists conducted in accordance with specified conducted an assessment of the map’s protocols and quality assurance accuracy during 2004. standards. Data obtained through this inventory are compatible, allowing for Products developed for ROMO are synthesis and analysis at broader levels described and presented in this report (http://www1.nature.nps.gov/protectingr and are stored on the accompanying estoring/IM/resourceinventories.cfm). DVD, these include:

To effectively classify and map the wide • A Final Report that includes a range of vegetation at ROMO required a vegetation key, accuracy assessment multi-year approach and consisted of information, and a map unit visual several linked phases: (1) vegetation guide; classification using the National • A Spatial Database containing digital Vegetation Classification (NVC), (2) vegetation map, plots, accuracy digital vegetation map production, and assessment, and flight line index (3) map accuracy assessment. To layers; classify the vegetation, we sampled 623 • Digital Photos of each vegetation representative plots located throughout type along with representative the 267,000-acre (108,000 ha) park ground photos and miscellaneous during the summer of 2003. Analysis of Park views; the plot data using ordination and • Field key for association clustering techniques produced 172 identification and a list of distinct plant associations, 19 of which associations present in the mapping were newly described. These include 14 area; NVC associations, and 5 local types. • Federal Geographic Data Committee-compliant metadata for To produce the digital map, we used a all spatial database coverages and combination of 2001 1:12,000-scale true field data. color aerial photography, 2001 1:40,000- scale true color ortho-rectified imagery In addition, ROMO and the USGS CBI reproduced at 1:12,000-scale, and 3 both received copies of: years of ground-truthing to interpret the complex patterns of vegetation and • 9x9 inch Aerial Photos;

1 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

• Uncompressed individual Digital to NatureServe’s website: Orthophotos and a compressed http://www.natureserve.org. Colorado MrSid© compilation of Digital Natural Heritage Program and BOR Orthophotos; have numerous services and programs • Digital data files and hard copy data and may be visited at sheets of the observation points, http://www.cnhp.colostate.edu and vegetation field plots, and accuracy http://www.usbr.gov. assessment sites; • Hardcopy, paper vegetation maps.

The DVD attached to this report contains text and metadata files, keys, lists, field data, spatial data, the vegetation map, graphics, and ground photos. The USGS will post this project on its website: http://biology.usgs.gov/npsveg/index.ht ml

For more information on the NVCS and NVC associations in the U.S. please go

Horseshoe Park

2 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

INTRODUCTION

Background protocols facilitate effective resource stewardship by ensuring compatibility USGS-NPS Park Vegetation and widespread use of the information Mapping Program throughout the NPS as well as by other federal and state agencies. These In 1994, the U.S. Geological Survey vegetation maps and associated (USGS) and National Park Service information support a wide variety of (NPS) formed a partnership to map resource assessment, park management, National Parks in the United States using and planning needs, and provide a the National Vegetation Classification structure for framing and answering (NVC). The goals of the USGS-NPS critical scientific questions about Vegetation Mapping Program are to vegetation communities and their provide baseline ecological data for park relationship to environmental processes resource managers, create data in a across the landscape. regional and national context, and provide opportunities for future The NVC has primarily been developed inventory, monitoring, and research and implemented by The Nature activities (FGDC 1997, Grossman et al. Conservancy (TNC) and the network of 1998, Natural Heritage Programs over the past http://biology.usgs.gov/npsveg/index.ht twenty years (Grossman et al. 1998). ml). Currently the NVC is maintained and updated by NatureServe. Additional Central to fulfilling the goals of this support has come from federal agencies, national program is the use of the the Federal Geographic Data Committee National Vegetation Classification (FGDC), and the Ecological Society of (NVC) as the standard vegetation America. Refinements to the classification. This classification: classification occur in the application process, leading to ongoing proposed • is based upon current vegetation; revisions that are reviewed both locally • uses a systematic approach to and nationally. NatureServe has made classify a continuum; available a 2-volume publication • emphasizes natural and existing presenting the standardized vegetation; classification. This document provides a • uses a combined physiognomic- thorough introduction to the floristic hierarchy; classification, its structure, and the list of vegetation types found across the United • identifies vegetation units based States as of April 1997 (Grossman et al. on both qualitative and 1998). This publication can be found on quantitative data; the Internet at: • is appropriate for mapping at http://www.natureserve.org/publications/ multiple scales. library.jsp.

NatureServe has since superseded The use of standard national vegetation Volume II (the classification listing) classification and mapping with an oline database server that provides regular updates to ecological

3 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park ommunities in the United States and to undertake the mapping portion of this Canada. NatureServe Explorer®, can project. At this time Colorado Natural also be found on the Internet at: Heritage Program and NatureServe were http://www.natureserve.org/explorer. also contracted to conduct both stages of fieldwork (initial classification and Rocky Mountain National accuracy assessment) and classification Park Vegetation Mapping stages. Project Colorado Natural Heritage Program,

NatureServe, BOR RSGIS, and the Park The specific decision to map the ultimately formed a four-part vegetation vegetation at ROMO as part of the U.S. team each responsible for a specific Vegetation Mapping Program was made portion of the project as outlined by CBI in response to the NPS Natural (Appendix A). Colorado Natural Resources Inventory and Monitoring Heritage Program and NatureServe Guidelines issued in 1992. Under these became primarily responsible for guidelines, ROMO was viewed as a top- collecting standardized field samples and priority Park based on its need for the using them to classify ROMO’s program’s vegetation map products. vegetation types and also to provide data Driving this need was the Park’s for an accuracy assessment on the final inability to spatially analyze the vegetation map. RSGIS took on the role vegetation at a fine enough scale to of the mapping team responsible for accurately predict various management aerial photo interpretation and creation issues. Central to their concerns were of a digital vegetation map. Finally, the need for modeling the spread and ROMO staff provided logistical and intensity of fire and the management of a technical support, helped coordinate number of native and non-native flora fieldwork, and reviewed and evaluated and fauna. draft data.

In 2000 the USGS Center for Biological As a team, our objectives were to Informatics (CBI) kicked-off this project produce final products consistent with by asking the U.S. Bureau of the national program’s mandates. These Reclamation’s Remote Sensing and included the following: Geographic Information Group (RSGIG)

4 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Spatial Data

ƒ Aerial photography ƒ Map classification ƒ Map classification description and key ƒ Spatial database of vegetation communities ƒ Hardcopy maps of vegetation communities ƒ Metadata for spatial databases ƒ Complete accuracy assessment of spatial data

Vegetation Information

ƒ Vegetation classification ƒ Dichotomous field key of vegetation classes ƒ Formal description for each vegetation class ƒ Ground photos of vegetation classes ƒ Field data in database format

Previous Vegetation Maps The previous vegetation map was a Only one previous effort has been compilation of several maps for the Park completed for a vegetation map of the and the various State and Federal lands entire Park. The existing vegetation map adjacent to the Park. Because this is a was developed in the late 1980’s. The digital compilation one does find differences in resolution, both spatially artifacts such as sliver polygons that are and in the classification for the mapped 0.001 hectare in size. Summary area within the Park boundary is detailed statistics were difficult to determine and in Table 1. The current map is information about how the map was considerably more detailed in both made could not be found. No attempt polygon size and number of map units. was made to develop metadata for this In addition there is considerably more previoius project beyond using it for detailed floristic description of the comparison. vegetation communities.

Table 1. Summary statistical comparison of current map effort to existing vegetation map.

Old Vegetation Current Vegetation Statistic Map Map Number of Polygons 7,141 22,065 Mean Polygon size (acres) 37.7 13.1 Range in Polygon Size (acres) 0.001-26,298 0.03-3,254 Number of Map Units 33 46

5 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Scope of Work classification, field key, and local association descriptions were completed Vegetation at ROMO was mapped and during 2003. Planning for the accuracy classified by a consortium that included assessment was conducted during the the U.S. Bureau of Reclamation, spring of 2004 and the AA data were Colorado Natural Heritage Program and collected over the summer of 2004. The Natureserve. The protocols and assessment of the map accuracy was standards used are described in the completed during the winter of 2004 and NPS/BRD program documents. BOR spring of 2005. was contracted by USGS-BRD in 2001 to have 652 square miles in an around Vegetation mapping required the close ROMO flown and photographed to cooperation of the three primary obtain 1:12,000 true color stereo-pair contractors for this project and aerial photographs and 1:12,000 scale communication between all three and the true color digital ortho-photography (1 co-project managers (Jeff Connor and meter pixels) using the USGS DOQQ Ron Thomas - NPS ROMO) was specifications. instrumental to this very large effort.

Vegetation mapping for ROMO encompassed both the executive boundary of ROMO, a 1 mile environ radius or buffer to the north, west and south and approximately 4 miles from the eastern most extent of the Park boundary. The 1 mile buffer is standard for most park mapping projects. The extended buffer to the east of the Park was to include data available for broader management concerns that included the extensive urban interface. Wildfire concerns for this area largely drove the decision to include this large buffer (Figure 1).

The project was initiated in the Fall of 2001 with the acquisition of the aerial imagery and completion of scoping meetings. Project planning and logistics were completed during the winter and spring of 2001- 2002, and vegetation data were collected during the summer of 2002. The vegetation

6 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 1. Vegetation mapping – project and map boundaries.

7 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Introduction to the National Vegetation Classification (NVC) and System (NVCS)

The Vegetation Mapping Program uses The US NVC is part of the International the US National Vegetation Vegetation Classification System (IVC) Classification (NVC) as the standard to which currently includes the USA, identify and describe vegetation types Canada, and several , Central within the map boundaries. The and South American countries. Its National Vegetation Classification application is rapidly expanding and (NVC) was begun in the early 1990’s by may soon include other countries as ecologists in the Science Division of The well. Nature Conservancy and state Natural Heritage Programs and Conservation Data Centers in collaboration with The NVC uses a hierarchical system of 7 partners from the academic, levels; the lower levels are nested into conservation, and government sectors the higher levels. The two lowest levels and is now managed and maintained by (most specific), Alliance and NatureServe. This classification was Association, are based entirely on the designed to allow description of plant floristics, while the upper five levels are assemblages based on existing based on physiognomy (structural and vegetation rather than on potential morphological characteristics of the natural vegetation, climax vegetation, or vegetation type e.g. forest, grassland, physical habitats. The classification evergreen, deciduous, broadleaved , currently includes more than 5000 needle-leaved), natural and cultural vegetation associations and 1,800 characteristics, and flood regime. Table Alliances, and has been adopted by the 2 identifies the 7 levels of the NVC and Federal Geographic Data Committee for depicts their placement in the use by all U.S. federal agencies. hierarchical relationship (Maybury 1999).

Table 2. Summary of the National Vegetation Classification System Hierarchal Approach (Maybury 1999).

LEVEL PRIMARY BASIS FOR CLASSIFICATION EXAMPLE

Class Structure of vegetation Woodland Subclass Leaf phenology Evergreen Woodland Group Leaf types, corresponding to climate Temperate or Subpolar Needle- Leaved Evergreen Woodland Subgroup Relative human impact (natural/semi-natural, or Natural/Semi-natural cultural) Formation Additional physiognomic and environmental Saturated Temperate or Subpolar factors, including hydrology Needle-Leaved Evergreen Woodland Alliance Dominant/diagnostic species of the uppermost Longleaf Pine -- (Slash Pine, Pond or dominant stratum Pine) Saturated Woodland Alliance Association Additional dominant/diagnostic species from any Longleaf Pine / Little Gallberry / strata Carolina Wiregrass Woodland

8 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Alliances and Associations are based on • Parentheses around a species both the dominant (greatest canopy name indicates the species is less cover) species in the upper strata of a consistently found either in all associations of an alliance, or in stand as well as on diagnostic species all occurrences of an association. (those consistently found in some types but not others). Associations are the • Association names include the most specific classification and are dominant species of the hierarchically subsumed in the significant strata, followed by the Alliances. Each Association is included class in which they are classified in only one Alliance, while each (e.g., "Forest," "Woodland," or Alliance typically includes many "Herbaceous Vegetation"). Associations. Alliance names are • Alliance names also include the generally based on the class in which they are classified dominant/diagnostic species in the (e.g., "Forest," "Woodland," uppermost stratum of the vegetation, "Herbaceous"), but are followed though up to four species may be used if by the word "Alliance" to necessary to define the type. distinguish them from Associations define a distinct plant Associations. composition which repeats across the landscape and are generally named using Examples of alliance names from both the dominant species in the ROMO: uppermost stratum of the vegetation and one or more dominant species in lower • Pseudotsuga menziesii Forest strata, or a diagnostic species in any Alliance stratum. A table listing all the dominant • Pinus flexilis - Populus species is included in Appendix K. The tremuloides Forest Alliance species nomenclature for all Alliances and Associations follows that of Kartesz Examples of association names from (1999). Documentation from ROMO: NatureServe (2005) describes the naming and syntax for all NVC names: • Abies lasiocarpa / Mertensia ciliata Forest

• Artemisia tridentata ssp. • A hyphen ("-") separates names of species occurring in vaseyana - (Purshia tridentata) / the same stratum. Muhlenbergia montana - (Hesperostipa comata ssp. • A slash ("/") separates names of species occurring in comata) Shrubland different strata. • Carex rupestris - Trifolium dasyphyllum Herbaceous • Species that occur in the uppermost stratum are listed Vegetation first, followed successively by those in lower strata. For more information on the NVC see Grossman et al. (1998). • Order of species names generally reflects decreasing levels of dominance, constancy, or indicator value.

9 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

In addition to the NVC, NatureServe has likely to occur. Because the structure of created standardized Ecological Systems the NVC is hierarchical, each association Classification for describing sites based occurs in only one alliance. An on both the vegetation and the ecological association may occur in any number of processes that drive them. Ecological Ecological Systems, limited only by the systems are mid-scale biological range of ecological settings in which that communities that occur in similar Association occurs. Ecological Systems physical environments and are are much like the map units used for the influenced by similar dynamic map legend; they are a broader scale ecological processes, such as fire or concept that embodies the concepts of flooding. They are not conceptually a several highly specific Associations that unit within the NVC and do not occupy a might be found in a particular setting. place in the NVC hierarchy. However, A field key for ecological systems is within each Ecological System resides a included in Appendix F. specific list of NVC associations that are

10 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Introduction to Natural Heritage Program Methodology and Element Ranking

The Colorado Natural Heritage Program imperilment status provides guidance for (CNHP) is a member of the NatureServe where Natural Heritage Programs should Network of Natural Heritage Programs focus their information-gathering and Conservation Data Centers. The activities and provides data users with a Natural Heritage Programs (and concise and meaningful tool for conservation data centers) are located in decision-making. all the States and Canadian Provinces. Each Program serves as that state's To determine the status of an element biological diversity data center, within Colorado, CNHP gathers gathering information and field information on plants, animals, and plant observations to help develop national communities. Each of these elements of and statewide conservation priorities. natural diversity is assigned a rank that indicates its relative degree of The multi-disciplinary team of scientists, imperilment on a five-point scale (1 = planners, and information managers at critically imperiled, 5 = demonstrably the Heritage Programs use a secure). The criteria used to define the standardized methodology to gather element imperilment rank are number of information on the rare, threatened, and occurrences, size of population, and endangered species and significant plant quality of population. The primary communities that occur in each state. criterion is the number of occurrences The species and plant communities each (in other words, the number of known Program maintains data for are referred distinct localities or populations). This to as “elements of natural diversity” or factor is weighted more heavily than simply “elements”. Life history, status, other factors because an element found and locational data are regularly updated in one place is more imperiled than in a comprehensive shared data system. something found in twenty-one places. Sources of element data include Also of importance are the size of the published and unpublished literature, geographic range, the number of museum and herbaria labels, and field individuals, the trends in both population surveys conducted by knowledgeable and distribution, identifiable threats, and naturalists, experts, agency personnel, the number of protected occurrences. and our own staff of botanists, ecologists, and zoologists. Element imperilment ranks are assigned both in terms of the element's degree of The Natural Heritage Ranking imperilment within Colorado (its State- System rank or S-rank) and the element's imperilment over its entire range (its The cornerstone of Natural Heritage Global-rank or G-rank). Taken together, methodology is the use of a standardized these two ranks indicate the degree of element imperilment ranking system. imperilment of an element. For Ranking species and ecological example, the lynx, which is thought to communities according to their be secure in northern but

11 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park is known from less than five current whether more active tracking is locations in Colorado, is ranked G5 S1 warranted. A complete description of (globally-secure, but critically imperiled each of the Natural Heritage ranks is in this state). The Rocky Mountain provided in Table 3. Columbine, which is known only in Colorado from about 30 locations, is This single rank system works readily ranked a G3 S3 (vulnerable both in the for all elements except migratory animal state and globally, since it only occurs in species. Those animals that migrate may Colorado and then in small numbers). spend only a portion of their life cycles Further, a tiger beetle that is only known within the state. In these cases, it is from one location in the world at the necessary to distinguish between Great Sand Dunes National Park and breeding, non-breeding, and resident Preserve is ranked G1 S1 (critically species. As noted in Table 3, ranks imperiled both in the state and globally, followed by a "B,” for example S1B, because it exists in a single location). indicate that the rank applies only to the CNHP actively collects, maps, and status of breeding occurrences. electronically processes specific Similarly, ranks followed by an "N” occurrence information for animal and refer to non-breeding status, typically plant species considered extremely during migration and winter. Elements imperiled to vulnerable in the state (S1 - without this notation are believed to be S3). Certain elements are “watchlisted,” year-round residents within the state. meaning that specific occurrence data are periodically analyzed to determine

Table 3. Definition of Natural Heritage Imperilment Ranks.

Note: Where two numbers appear in a state or global rank (for example, S2S3), the actual rank of the element is uncertain, but falls within the stated range. G/S1 Critically Imperiled globally/state because of rarity (5 or fewer occurrences in the world/state; or 1,000 or fewer individuals), or because some factor of its biology makes it especially vulnerable to extinction. G/S2 Imperiled globally/state because of rarity (6 to 20 occurrences, or 1,000 to 3,000 individuals), or because other factors demonstrably make it very vulnerable to extinction throughout its range. G/S3 Vulnerable through its range or found locally in a restricted range (21 to 100 occurrences, or 3,000 to 10,000 individuals). G/S4 Apparently Secure globally/state, though it may be quite rare in parts of its range, especially at the periphery. Usually more than 100 occurrences and 10,000 individuals. G/S5 Demonstrably Secure globally/state, though it may be quite rare in parts of its range, especially at the periphery. G/SX Presumed Extinct globally, or extirpated within the state.

G#? Indicates uncertainty about an assigned global rank. G/SU Unable to assign rank due to lack of available information.

GQ Indicates uncertainty about taxonomic status.

G/SH Historically known, but usually not verified for an extended period of time. G#T# Trinomial rank (T) is used for subspecies or varieties. These taxa are ranked on the same criteria as

12 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

G1-G5. S#B Refers to the breeding season imperilment of elements that are not residents.

S#N Refers to the non-breeding season imperilment of elements that are not permanent residents. Where no consistent location can be discerned for migrants or non-breeding populations, a rank of SZN is used. SZ Migrant whose occurrences are too irregular, transitory, and/or dispersed to be reliably identified, mapped, and protected. SA Accidental in the state.

SR Reported to occur in the state but unverified.

S? Unranked. Some evidence that species may be imperiled, but awaiting formal rarity ranking.

Legal Designations for Rare an element occurrence rank (EO-Rank) Species is assigned according to the size, ecological quality and landscape context Natural Heritage imperilment ranks of the occurrences whenever sufficient should not be interpreted as legal information is available. This ranking designations. Although most species system is designed to indicate which protected under state or federal occurrences are the healthiest and endangered species laws are extremely ecologically the most viable, thus rare, not all rare species receive legal focusing conservation efforts where they protection. Legal status is designated by will be most successful. The EO-Rank either the U.S. Fish and Wildlife Service is based on three factors: under the Endangered Species Act or by the Colorado Division of Wildlife under Size – a measure of the area or Colorado Statutes 33-2-105 Article 2. In abundance of the element’s occurrence. addition, the U.S. Forest Service Takes into account factors such as area recognizes some species as “Sensitive,” of occupancy, population abundance, as does the Bureau of Land population density, population Management. fluctuation, and minimum dynamic area (which is the area needed to ensure Element Occurrences and their survival or re-establishment of an Ranking element after natural disturbance). This factor for an occurrence is evaluated Actual locations of elements, whether relative to other known, and/or presumed they are single organisms, populations, viable, examples. or plant communities, are referred to as element occurrences. The element Condition/Quality – an integrated occurrence is considered the most measure of the composition, structure, fundamental unit of conservation interest and biotic interactions that characterize and is at the heart of the Natural the occurrence. This includes measures Heritage Methodology. To prioritize such as reproduction, age structure, element occurrences for a given species, biological composition (such as the presence of exotic versus native species),

13 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park structure (for example, canopy, species having access to habitats and understory, and ground cover in a forest resources needed for life cycle community), and biotic interactions completion, fragmentation of ecological (such as levels of competition, predation, communities and systems, and the ability and disease). of the species to respond to environmental change through dispersal, Landscape Context – an integrated migration, or re-colonization. measure of two factors: the dominant environmental regimes and processes Each of these factors is rated on a scale that establish and maintain the element, of A through D, with A representing an and connectivity. Dominant excellent rank and D representing a poor environmental regimes and processes rank. These ranks for each factor are include herbivory, hydrologic and water then averaged to determine an chemistry regimes (surface and appropriate EO-Rank for the occurrence. groundwater), geomorphic processes, If not enough information is available to climatic regimes (temperature and rank an element occurrence, an EO-Rank precipitation), fire regimes, and many of E is assigned. EO-Ranks and their kinds of natural disturbances. definitions are summarized in Table 4. Connectivity includes such factors as a

Table 4. Element Occurrence Ranks and their Definitions

A Excellent viability. B Good viability. C Fair viability. D Poor viability. H Historic: known from historical record, but not verified for an extended period of time. X Extirpated (extinct within the state). E Extant: the occurrence does exist but not enough information is available to rank. F Failed to find: the occurrence could not be relocated.

14 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

PROJECT AREA

Location and Regional divide at 10,790 ft (3289 m). Setting (http://www.nps.gov/romo/pphtml/natur alfeatures.html). The Park is divided ROMO lies in north central Colorado on unequally into eastern (~60%) and the Eastern slope (Figure 2). The eastern western portions (~40%) by the entrance is situated adjacent to Estes continental divide which runs roughly Park and is accessed by either State from the southeast to the northwest Highway 36 from Boulder or State (Figure 3). The Park also lies within Highway 34 from Loveland / Fort Grand, Larimer and Boulder counties. Collins. The western entrance is Surrounding the Park are the Routt, adjacent to Grand Lake and is accessed Roosevelt and Arapahoe National from State Highway 34. State Highway Forests, the Colorado State Forest State 34 traverses the Park and is commonly Park and the Comanche Peak and Indian known as “”. It is the Peaks Wilderness areas. A populated / highest continually paved highway in the urban interface occurs all along the United States and peaks at just over eastern edge of the Park and along the 12,100 feet above sea level. Rocky southwest corner in the Grand Lake and Mountain is the highest national park in area. The Park boundary is the US, with elevations from 7860' to contiguous with both U. S. Forest 14,259' More than one-fourth of the Park Service lands (67%) and private lands is above tree line (11,200-11,500'). State (33%). Highway 34 crosses the continental

15 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 2. Location map – ROMO.

16 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 3. Detail map - ROMO

17 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Climate and Weather (http://instaar.colorado.edu/research/high lights.html#hydrology). Temperature is The climate in and around ROMO is also largely controlled by elevation. described as temperate semi arid steppe Temperatures at higher elevations will (Bailey R.G. 1995). The primary tend to fluctuate wildly as a result of controlling factor to climate is the north rapid gain and loss through a thinner – south orientation of the Rocky atmosphere (Kershaw et al. 1998). Mountains and its elevation gradient. Given the topographic variability Weather arriving from the west tends to throughout the Park, any meaningful leave its moisture on the western slopes; climatic information for a given locality hence the eastern slopes are typically is difficult (Hess and Alexander 1986). drier and warmer. This rain shadow Nonetheless, the Institute of and effect is wide and is exemplified Alpine Research (INSTARR) has in the Park by the mean annual collected climatological data for four precipitation of both the Pinus contorta sites in the area dating back to 1952. and Picea engelmannii- Abies lasiocarpa These data show a gradient in mean zones on the east and west sides of the annual temperature from 8.3°C at 2195 continental divide. In both cases the m to –3.3° C at 3,750 m. The mean annual precipitation is up to 5 elevational cooling effect equates to 7.5° inches more on the west than on the east C per 1,000 m of elevation (Peet 1981). side. (Haeffner 1971; Marr 1968a, 1968b). Winds can often have a severe and lasting impact on the vegetation. The Precipitation varies considerably and is alpine Krummholz owes its very primarily controlled by elevation. existence not only to the cold but also Precipitation at higher elevations usually the persistent winter winds and short falls as snow. The winter precipitation is growing season found in this zone. typically more predictable than summer These winds can be exceptionally strong as cyclonic storms tend to dominate during the spring and can alter the (Peet 1981). Lower elevations may landscape by creating large blowdowns receive 10 to 20 inches of rainfall while in forests (Peet 1981). These winds the peaks may receive twice that much. often exceed hurricane force, with Because of the proximity to the Gulf of speeds over 74 miles per hour (119 Mexico and the Gulf of California, the kilometers) in some areas. Anecdotal summer precipitation is monsoonal – accounts tell of a wind anemometer self that is, it arrives regularly and destructing after passing a recorded sometimes violently in the afternoons. speed of 200 miles per hour (Jeff The mountains tend to generate Connor– pers. comm.). thundershowers and thunderstorms, which then move elsewhere. Although Topography precipitation arrives year around, the annual hydrologic cycle is dominated by ROMO lies on the eastern edge of the wintertime accumulation snow and the Rocky Mountains and abuts the Great melting of the snowpack in spring Plains that extends eastward. The Park is located approximately 60 miles north of

18 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Denver, Colorado. The project area around it are derived from the boundary includes elevations from 2285 glacial deposits from this valley. m at Estes Park to 4,345 m at Longs Peak. The Park is divided in two by the Two prominent drainages exit the Park continental divide. There are a number to the northeast. These are the Fall of cirque glaciers, extensive alpine River and the Thompson River tundra, a variety of glacial landforms, drainages. Both these have large glacial lakes and moraines, cirques and moraine deposits forming Horseshoe talus slopes, patterned ground, and Park and Moraine Park respectively. To permafrost. The mapping area is the southeast is yet another important bounded on the west by the Never drainage, the North St. Vrain. Copeland Summer Range and on the east by the Moraine is formed from this valley’s hogbacks of the . outwash. To the southwest one finds the The range trends north to south with a headwaters of the large valley running north to south on flowing south through the Kawuneeche the west side of the Park (, into Shadow Mountain Lake, Valley)(Figure 4.) Grand Lake and the then into Lake Granby and then exiting to the southwest through Granby Dam.

Colorado River Headwaters

19 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 4. Topography of Rocky National Park end environs.

Geology subjected to tremendous pressures from the accumulated weight and buckled and The geology of ROMO dates back to the twisted and eventually were transformed Precambrian and includes several into metamorphic rocks. These episodes of mountain building and metamorphics are primarily gneisses with erosion, the evidence of which may be lesser amounts of schist and quartzites found within the Park. The Quaternary interlaced with material from the geological history is the most evident underlying magma chamber and lava feature within the Park. flows. These dark gray contorted bands of rock can be seen today along Trail Precambrian Ridge Road on the west side of the Park. Some time over two billion years ago the In fact, the majority of the exposed entire Rocky Mountain area was under geology within ROMO is metamorphic water. Sediments from surrounding areas and igneous rocks from the Precambrian. formed deposits tens of thousands of feet Approximately one billion years ago the deep. It is believed that the sediments underlying molten magma chamber began collected within a geosyncline created by to rise and eventually cooled to form the the continental crust riding over oceanic pink granites seen on the eastern side of crust. These sediments where then the Park (Silver Plume Granite).

20 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Paleozoic years ago. Most of these volcanic The next 500 million years or so were sediments have been eroded away marked by an extensive period of erosion however there are traces of this event in and the area was again covered by an the northwest corner of the Park at inland sea. There is no evidence of , Lava Cliffs and marine deposits during this period within on Trail Ridge Road. the Park (Middle Paleozoic). This was then followed by a new mountain Quaternary building event to form what we know as The end of Tertiary period saw the Rocky the ancestral Rockies. These Rockies Mountains with considerable elevation where thought to be a vast island with and V-shaped valleys. Then the ice age 2000 foot peaks about 100 miles to sea. began. Much of the current topography These ancestral Rockies eventually within ROMO is a direct result of the eroded to form the “Fountain Formation” sculpting and deposition of mountain which is a very distinctive red outcrop, glaciers. The Rocky Mountains are far steeply tilted and runs along the eastern enough south to have escaped the edge of the present mountain range from continental glaciation experienced in to South of Colorado Springs. northern North America but still experienced considerable mountain Mesozoic glaciation. There were a number of The ancestral Rockies eroded during the glaciations in the Rocky Mountains late Paleozoic to early Mesozoic (325 – during the Quaternary, however evidence 245 mya). The Mesozoic period was a for the earlier episodes is sparse. time of extensive marine deposition and is best known as the age of . It is generally believed that there where at Mesozoic deposits are poorly represented least four glaciations of which only the within the Park with only a few areas of last two have left any evidence within the Pierre shale in the North West section of Park. Previous glacial deposits (Pre-Bull the Park in the Never Summer Range and Lake) were eliminated by the actions of in the vicinity of Lake of the Clouds. the Pinedale and Bull Lake glaciations. In the high country the mountains were Cenozoic scraped away forming broad U-shaped As the Mesozoic came to an end the valleys that are typical of glaciated modern day Rocky Mountains began to valleys. Towards the lower elevations are appear. This episode is known as the the moraine deposits. “The oldest is Laramide Orogeny and is associated not represented by a moraine about three only with uplift but also volcanic activity miles west of Estes Park where the Big and erosion of the marine sediments. At Thompson River traverses a wide U- this time, much of the earlier Dakota shaped valley before entering its narrow, sandstone and sediments unglaciated canyon” (Chronic and where eroded away exposing the Chronic 1972). Others include the Precambrian granites and schists. A Aspen-Gold campground, Horseshoe secondary uplift beginning in the Park and Moraine Park. A terminal Oligocene pushed the Rockies back up. moraine also formed Grand Lake. Most The Never Summer Range experienced of the lakes within the Park are a result of extensive volcanism about 25 million

21 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park glacial action, whether by erosion “The formation of the soils of ROMO (cirques) or moraine deposition. has been strongly influenced by landform and climate. There is a wide Soils range in soil properties from the warmer and drier valleys to the high Soils play an important role in elevation tundra. Soils of the low determining the types of vegetation that elevation valleys are generally very might inhabit a site. The controlling deep, loamy, and, particularly on the factors in young soils derived from east side of the Park, have dark-colored Precambrian granites and schists is often surface horizons. In the floodplains texture and depth. Both of these factors, they are poorly or very poorly drained in turn, will affect water availability with stratified textures. On stream (Smith 1985). When the soil develops a terraces they are well drained. They significant horizon and the texture is fine, formed in alluvium from the nearby grasses and forbs can form a dense sod mountains. Soils of the glacial that inhibits tree generation (Peet 1988). moraines are very deep, well or In other areas that have poor drainage somewhat excessively drained, and wetland vegetation will appear. These loamy or sandy with a high content of characteristics can aid in the rock fragments. They formed in till photointerpretive process. The NRCS derived mainly from granite, gneiss, has produced a soil map for the Park and and schist. Soils of the Subalpine surrounding areas. The soil types mountain slopes are generally well or descriptions are as follows: somewhat excessively drained, loamy with a high content of rock fragments, and have light-colored surface horizons. Depth to the underlying bedrock ranges from shallow to very deep. Typically soil reaction becomes more acid with increasing elevation, as the climate becomes cooler and moister. These soils formed mainly in material weathered from granite, gneiss, and schist. Soils of the alpine mountains and ridges are generally well drained, loamy with a high content of rock fragments, strongly acid, and have dark-colored surface horizons. These soils formed mainly in material weathered from granite, gneiss, and schist. Poorly drained soils are common in landscape depressions and drainageways.” (Neve, L.A. 2000)

22 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Wildlife release of carbon and other greenhouse gases (Hauer et al. 1997). The Rocky Mountains have a wide variety of wildlife although the southern Precipitation typically arrives from the rockies are depauperate in large Pacific or Gulf of Mexico and falls as carnivores such as the North American winter snow and summer thunderstorms. Wolf (Canis lupus) and Grizzly Bear The winter snowfall normally is regional (Ursus arctos horribilis). Within the and widespread while the summer Park, the large carnivores are represented thunderstorms typically are local and primarily by the Mountain Lion (Felis violent producing episodic flows in concolor), perhaps the Wolverine (Gulo individual drainages. gulo) and occasionally the Lynx (Lynx canadensis)(Jeff Conner pers. Comm.). Winter snow may either accumulate on The large omnivores present include the the existing glaciers and permanent snow Black Bear (Ursus americanus), Coyote fields, sublimate or join the spring melt (Canis latrans), Bobcat (Lynx rufus) and in either surface or subsurface flow. the Red and Gray Fox (Vulpes vulpes and Winter precipitation may impact the Urocyon cinereoargentues). ecosystem with periglacial activities (freezing and thawing of soils leading to The greatest impact on vegetation and frost heave and gelifluction) and mass vegetation distribution lies with the movement such as avalanches. Within herbivores. The large herbivores in the the mapping area are 93 snow fields or Park are the Mule Deer (Odocoileus glaciers that range in size from 0.04 to hemionus) the Rocky Mountain Elk 13.3 hectares with a mean of 1.6 (Cervus elaphus) and the Big Horn Sheep hectares. . In this project a total of 145 (Ovis canadensis). There are also hectares of snowfields/glaciers numerous other small herbivores that weremapped. play an important role in vegetation patterns with their herbivory and seed The spring snowmelt flows down 29 km dispersal, however the large herbivores of canals () and 960 km of are of particular concern due to their tributaries and streams within the Park impact on aspen groves. Aspen makes boundaries into or through numerous up a small percentage of the Park but its lakes and ponds. There are 410 mapped importance vastly outweighs its presence. lakes or ponds that range in size from Aspen is a critically important habitat to 0.01 hectares to 14 hectares (mean size a number of species including elk. Elk 1hectare) for a sum of 428 hectares of forage heavily in these areas and, given water surface within Park boundaries. the very high number of Elk in the Park, Larger reservoirs exist outside the Park. adversely impact the viability of aspen These are on the northwest boundary of groves. the Park (Long Draw Reservoir), the central-eastern mapping area (Lake Hydrology Estes) and to the southwest (Shadow Mountain Lake and Lake Granby). The hydrology of ROMO may be Adjacent to Shadow Mountain Lake and considered in two forms: its arrival and Lake Granby is Grand Lake which is the its movement. Each of these will impact largest natural lake in Colorado. the soil, vegetation, erosion, and even the

23 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Anticipated long term hydrologic impacts changes affecting species composition; due to climate change include a “Rise in increased isolation of coldwater stream snow line in winter-spring, possible fish ( Hofmann et al. 1998. Fyfe and increases in snowfall, earlier snowmelt, Flato 1999, McCabe and Wolock 1999, more frequent rain on snow - changes in Leith and Whitfield 1998, Williams et al. seasonal streamflow, possible reductions 1996, Hauer et al. 1997).” (From in summer streamflow, reduced summer http://www.isse.ucar.edu/water_climate/h soil moisture and stream temperature tml_map.html#3”).

24 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Vegetation interest we also briefly describe the east- west slope effects and the extensive Overview wetland/riparian systems present within the Park. Within each recognized zone ROMO lies within the Dry Domain, exists a variety of NVC plant alliances, Temperate Steppe Division and Southern and even more associations (see Rocky Mountain Steppe – Open Appendix I). Woodland – Coniferous Forest – Alpine Meadow Province as described by Bailey The concept of “zones” is complicated (1995). This ecoregion is characterized and exacerbated by plants refusing to by dramatic vertical zonation of follow the paradigms of previous vegetation. This zonation is a investigators (Ramaley 1907, 1908, consequence of abrupt elevational Rydberg 1916, Daubenmire 1938, 1943, gradients between flatlands and Costello 1954 and Moir 1969). mountains. Topographic relief is quite Peculiar groupings of soil, aspect, dramatic and in a short distance one may elevation and other controlling factors see various life zones as described by may allow species associated with a Merriam and Steineger (1890). particular zone to appear where they ought not. Indeed, the distribution of ROMO has all but the lower two life vegetation within the Park may be linked zones present (Upper and Lower to various factors, all of which are Sonoran). The Pinyon-Juniper (Upper interrelated in one form or another. Sonoran) life zone begins a short distance These include dramatic topography and to the east of the Park and a few foothill elevation, a variety of soils, moisture, communities occur within the study area temperature, wind, aspect, and slopes. on warm aspects. The primary factors controlling vegetation distribution within the Front The concept of life zones or landscape Range are elevation, moisture, units has been addressed by numerous disturbance (Peet 1981), and soils (Peet researchers since the late 19th century. 1988). These include Merriam (1898), Gregg (1947), Daubenmire (1938) Rydberg (1916) and Marr (1967) amongst others. East/West Slope Effects More recently, Peet (1981, 1988), quite One important factor affecting the extensively described the vegetation distribution of vegetation in ROMO is the within the Park. All of these researchers continental divide, which dissects the recognize three major vegetation zones or Park from north to south into two unequal based upon their easily halves, the larger half (63%) being the recognizable major flora although they eastern side. According to Beidleman et have often named them differently and or al. (2000), the western slope (everything sub-divided them into sub-zones or west of the divide) receives more rain in regions. The three primary vegetation summer and more snow in winter, than zones described and mapped in this effort does the windier eastern slope are the alpine, Subalpine and montane (everything east of the divide). This is zones (Table 5). As a matter of local

25 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

due to the mountains blocking weather riparian areas are dominated by tall that comes predominantly from the west. willows. Though the western slope This phenomenon is typically referred to contains many of the same communities as orographic precipitation on the west as the eastern slope, Pinus contorta slope and the “rain shadow” effect (or (lodgepole pine) and Populus tremuloides more accurately, precipitation shadow) (aspen) are more common. Artemisia on the eastern slopes. Beidleman also tridentata (sagebrush) communities, mentions that the divide acts as a “barrier occur mostly on the east side although to the spread of some plant species,” there do occur patches within the Park in especially those from lower elevations. the south west. Outside the Park one Because of these factors, the east and finds more sagebrush to the southwest, west sides of the Park are floristically especially in the Granby Lake area. At different, with the greatest differences higher elevations in the Subalpine and occurring in the montane zone. alpine zones, environmental conditions created by the divide affect vegetation, In very general terms, montane forests on but the barrier to migration is less the eastern slope are dominated by Pinus pronounced since the habitat of these ponderosa (Ponderosa pine) on south species is continuous across the divide, facing slopes and by Pseudotsuga thus, the east/west slope effect is less menziesii (Douglas-fir) on north facing apparent the higher one climbs. slopes. Grasslands are common, and

Table 5. Life zones as described by Merriam and Steineger (1890) and their modern equivalent.

(from http://www.cpluhna.nau.edu/Biota/merriam.htm) “Elevations given above modified for the latitude at ROMO” (Beidleman et al. 2000) and from estimations from Peet (1981).

Merriam's Life Modern Vegetation Zones General Name Elevation

Zones 1891 Used During Range (feet)

Project Arctic-Alpine Alpine Tundra Alpine 10,900-14,255 Hudsonian Spruce-Fir or Subalpine Subalpine 9,000-10,900 Conifer Forest Canadian Mixed Conifer Forest Upper Montane 8,000-9500 Transition Ponderosa pine Forest Lower Montane 6000-8500 Upper Sonoran Pinyon-Juniper Woodland, Foothill n/a Semi-Arid Grasslands, Semi-Arid Scrub Lower Sonoran Mojave, Sonoran, or NA n/a

26 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Alpine Zone Alpine mountain crests Alpine rock slides Perhaps the harshest environment within the Alpine mountain seeps Park is the Alpine zone. Here one Alpine meadows encounters the greatest extremes in Alpine bogs temperature and insolation coupled with Snow drift formations thin, or very young, soils. The alpine zone Alpine lakes in the Rocky Mountains has been extensively studied and factors controlling Billings (1988) describes a vegetational plant distribution in this zone are fairly well gradient that goes from the windward to known. The primary controlling factors for leeward slope that includes seven very alpine vegetation distribution appear to be general plant communities along a transect. “topographic exposure and distribution of These include: snow and meltwater superimposed upon geological substratum patterns” (Billings Windward slope from –150 m to and Mooney 1968). Additional studies ridge top at 0 m – open rock report similar results (Isard 1986; Marr fellfield. 1961; Haase 1987, Willard 1979, Zwinger Upper lee slope from 0 m to 40 m: and Willard 1972); that is, snow cover and modified fellfield soil moisture control the distribution of Upper lee slope from 40 m to 60 m: plant communities. transitional from fellfield to early snowbed community Categorizing the alpine tundra has taken on Middle lee slope from 60 m to 90 m: numerous shapes and sizes. There are many early snowbed community authors that describe the alpine vegetation Middle lee slope from 90 m to 120 and categorize them differently. m: late snowbed community Daubenmire (1943) breaks this zone into Middle lee slope from 120 m to 150 three rough, non-defined edaphic categories m: moist meadow that include very young, almost nonexistent Lower lee slope from 150 m to 260 soils, an intermediate gravelly and boulder m: mostly late snowbed community category, and a relatively more mature Bottom of lee slope from 260 m to category of thin soils. Clements (1904), in 300 m: very wet meadow his study of the area, identified five formations within the alpine zone The youngest of Daubenmire’s categories which include: consists of the more stable “boulder fields”. Little vegetation is found here with Alpine meadow formation exception of crustose lichens and crevice Alpine bog formation plants such as Oxyria digyna, Aquilegia Alpine lake formation spp., Polemonium spp., Pentstemon Alpine mat formation fruticosus and Sibbaldia procumbens. Alpine rock formation Areas with gravelly soils (fellfields) have Rydberg (1914) divides the zone into eight very sparse vegetation characterized by mat categories: or cushion plants such as Silene acaulis,

27 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Dryas octopetala, Arenaria sajanesis, montane zone, often extends well into the Erigeron compositus, E. multiflorus, Luzula Subalpine and, conversely, Engelmann spicata, Paronychia spp., Phlox caespitosa spruce often extends well into the montane and Selaginella densa (Daubenmire 1943). zone (Rydberg 1915). Engelmann spruce is typically found on northern slopes in either Areas with some soil development, typically pure stands or mixed with Subalpine fir. lower in the alpine zone, may be covered Subalpine fir rarely forms homogenous with dense, low lying vegetation often stands and is usually found growing with referred to as “alpine meadow” or “alpine Engelmann spruce. Aspen will grow grassland”. Species density increases interspersed with Engelmann spruce or it dramatically and includes numerous may form large groves. In addition, aspen grasses, forbs and sedges, including Carex will grow on slopes, but can also be found spp., Kobresia myosuroides, Poa spp., in moist areas with rich soil. Lodgepole Phleum alpinum, Deschampsia caespitosa, pine and aspen both regenerate rapidly after Trisetum subspicatum, Agrostis spp., fire. When fire does invade the Subalpine Festuca spp., Polygonum viviparum, zone, and the fire is intense enough to create Potentilla spp., Sieversia turbinate, a “stand replacement,” it is often colonized Trifolium spp. and Pedicularis parryi by dense, monotypic stands of lodgepole (Daubenmire 1943). pine, aspen, and sometimes Douglas-fir. The colonizing species will typically not Subalpine Zone establish themselves near timberline, but rather, spruce and fir regenerate directly The Subalpine zone is characterized by the (Daubenmire 1943). two dominant species of Picea engelmannii (Engelmann spruce) and Abies lasiocarpa Subalpine understory is typically very (Subalpine fir), though other authors similar over the entire range and is often (Rydberg 1915) count Subalpine fir as a composed of continuous cover of either secondary species in this zone, and aspen as Vaccinium scoparium or V. myrtillus (Peet a primary. Pinus aristida (bristlecone pine) 1981, 1988). Few other species are found is also considered a secondary species in the understory, but some species that within this zone; however, it’s known sometimes occur are Arnica cordifolia, northern most extent is just south of Carex geyeri, Pyrola secunda, and ROMO, and is unlikely to be found within Polemonium delicatum (Daubenmire 1943, the Park. However, Pinus flexilis (limber Peet 1988). pine) occurs in similar habitats in the Park and extend to the northern Rocky Within each zone, vegetation variation is Mountains. At the upper elevational end of often a function of moisture; for example, this zone, the trees take on a stunted growth Hess and Alexander (1986) report that the form known as Krummholz due primarily to Abies lasiocarpa/Vaccinium scoparium extreme conditions of a short growing habitat type occupies mesic upland season, low temperatures and desiccating positions while Abies lasiocarpa / winds. The lower elevation boundary Calamagrostis canadensis is typically found between the montane Douglas-fir zone and on wetter sites and topographically lower. the Subalpine zone is somewhat amorphous as Douglas-fir, the primary tree in the

28 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

The Subalpine vegetation is not restricted to comata, and other grasses, form a only forest and woodland ecosystems. patchwork with the Ponderosa pine forests. Rydberg (1915) identifies five grassland - Douglas-fir forests tend to be dense, with a forb types within the Subalpine. These sparse understory, and are susceptible to types are as follows; 1) Lakes, Ponds, spruce budworm. Other common trees that Brooks and Swamps, 2) Meadows, 3) Dry often grow mixed with Douglas-fir include Valleys and Bench-lands, 4) Mountain lodgepole pine and aspen, which are often slopes, and 5) Hog-backs. Hess and early seral species (Beidleman 2000). Alexander (1986) also report extensive high elevation grasslands dominated by Festuca On the wetter, west side of the Park thurberi on Arapahoe National Forest, lodgepole pine stands are more prevalent, adjacent to the west of the Park. and have a more diverse understory of shrubs. Aspen groves are also more Montane Zones common in wet areas. Shrublands dominated by Artemisia tridentata ssp. The montane zones (upper and lower) in vaseyana (mountain sagebrush) and ROMO occur mostly on the and bitterbrush, which are less common on the west sides of the Park at elevations below eastern slope, occur in abundance on the 3000m (9,800 ft.), and are dominated by western slope. Lupinus spp. (lupine) and mixed to single species woodlands and Balsamorhiza sagittata (arrowleaf forests, alternating with shrublands and dry balsamroot), which don’t often grow on the grasslands on hillsides and ridges, and with east side, are common here (Beidleman riparian areas in valleys and gullies. 2000).

On the east side of the Park, the montane Peculiar to the montane zone on the east zone is dominated by open Ponderosa pine side of the Park is the distribution of forests on south facing slopes, and by mixed Douglas-fir. Typically one finds Douglas- conifer forests (generally dominated by fir on north facing slopes or other protected Douglas-fir) on the more northern aspects. microclimates. Recently, however, one Understory species in the Ponderosa pine finds that many Ponderosa pine stands have forests generally include shrubs such as a rather extensive sub-canopy of Douglas- Purshia tridentata (bitterbrush), and dry fir. It has been hypothesized by some grassland species. In the lower elevations (Dickman 1978, Peet 1981, Covington and of this zone the south facing slopes are often Moore 1992, Veblen 1998, Brown et al. dominated by shrubs and herbaceous 1999, Kaufmann et al. 2001) that this is a species mixed with Ponderosa pine and / or result of effective fire suppression exercised Juniper. The north facing slopes often are throughout the region during the later parts dominated by Ponderosa pine, Douglas-fir, of the 20th century. Juniper or a combination of these. Ponderosa pine forests in ROMO are Wetland and Riparian sometimes attacked by Rocky Mountain pine beetles and mistletoe. Ponderosa pine Wetlands and riparian areas are abundant can be an early seral species after throughout the Park at all elevations and disturbance. Dry grasslands, dominated by vary according to elevation and topography. Muhlenbergia montana, Hesperostipa In the alpine zone, wet meadows and bogs

29 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park contain a variety of sedges, rushes and In the montane zones, wetlands and riparian forbs, the most common being Carex areas can include rivers, lakes, reservoirs, scopulorum and Caltha leptosepala. Near and wide glacial valleys. Riparian treeline, wet areas are usually home to vegetation along rivers can be dominated by communities of short willows such as Salix Picea pungens (Colorado blue spruce), or planifolia and Salix brachycarpa, which by a mixture of tall willows, aspen and often grow mixed with Betula nana and alder. Lakes and reservoirs are often dwarfed spruce and fir. Wide riparian surrounded by wet meadows similar to valleys in the Subalpine zone also contain those found in the glacial valleys; these this mix of shrubs, and often have a diverse glacial valley and lakeside meadows are understory of forbs and graminoids. Creeks dominated by graminoid species such as and lakes are abundant in the Subalpine; Juncus balticus and Carex aquatilis, with wet meadows, and long, thin riparian strips, Carex utriculata in particularly wet areas. which often contain tall willows and aspens, Willows are often present in strips along can be found near these. streams in these meadows, and occasionally whole valleys can be filled with mixed willow stands.

30 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

METHODS • Transfer interpreted information to a digital spatial database and produce The methods to produce a vegetation map hard copy (paper) vegetation maps; of this magnitude involve a tremendous • Create digital vegetation coverages amount of interaction and communication including relevant attribute between numerous groups and individuals. information; All of these interactions are summed up in a • Produce Arc/Info export file of flow chart that describes the general task vegetation plot, observation point, assignments and relationships (Appendix and accuracy assessment locations; A). The general groups of tasks include • Provide an annotated list of planning meetings, collecting and analyzing representative field site existing data, development of the photographs/slides; classification, development of the sampling • Create a contingency table strategy, field work, data input and analysis, comparing the mapped classes with photointerpretation, cartography, and map the AA classes in order to determine validation and accuracy assessment. These map accuracy; tasks necessarily interact with one another • Provide any ancillary digital files throughout the entire process. developed during the mapping process; Planning and Scoping • Document and record digital FGDC compliant metadata files (*.html) for This project incorporated the combined all created spatial data; expertise and oversight of several • Produce the final report and CD- organizations. Oversight and programmatic ROM describing procedures used in considerations were managed by the Center preparing all products; for Biological Informatics (CBI) of the USGS/BRD. NPS and ROMO personnel NatureServe Responsibilities and provided additional guidance on specific Deliverables: Park needs. The technical mapping portion was contracted to the BOR RSGIG in • Assist in project scoping and Denver, CO. CNHP was contracted planning activities; separately to collect, analyze, and write-up • Develop a preliminary vegetation the requisite plant association data and classification for the study area from conduct fieldwork to support the accuracy secondary sources; assessment (AA). NatureServe was • Design a sampling strategy using contracted to support the field surveys and Gradsect design; data analysis. The specific technical • Assist in training field crews in responsibilities and deliverables for the standard NPS vegetation sampling mapping portion included the following: methods for classification and accuracy assessment; BOR Responsibilities and • Be available for consultation in data Deliverables: management and vegetation classification; • Assist in developing mapping units; • Interpret aerial photographs;

31 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

• Support photo-interpretation by Deliverables spending time in field with photo- • Data in digital form from Vegetation interpreters and being available of Plots and Accuracy Assessment consultation; Points; • Review and finalize draft • Classification of ROMO vegetation; classification, local community • Local vegetation descriptions; descriptions, and field key to • Key to the Vegetation of ROMO; community types; • Final Report. • Complete the global descriptions that characterize the vegetation types Scoping Meetings throughout its distribution; • Conduct targeted sampling to The project participants met on several resolve classification issues and occasions over the course of the project to sample undersampled or unsampled discuss project progress and develop the vegetation associations; scope of activities that needed to be • Assist in analysis of Accuracy completed over the near-term. These Assessment data; informally structured “scoping” meetings • Assist in preparation of final report. were held at least once per year, often more frequently. The following is a summary of CNHP Responsibilities and several of the most pertinent of these. Deliverables: Initial Planning Meetings: Responsibilities Several initial planning meetings were held • Research existing vegetation data prior to the start of any field work. These from ROMO; • Collect 500+ were intended to provide the forum for Vegetation Plots; discussing various logistical issues and • Work closely with BOR photo determining the scope and scale of various interpreters to assure a common project aspects. Questions of whether and understanding of the classification of how to permanently mark the plots, the vegetation in the Park; availability of crew housing, backcountry • Enter and analyze data from camping, land access, the content of the Vegetation Plots to produce a final preliminary association list, the content of classification for mapping; the map unit list, and other questions were • Collect, and enter data for, 1200+ discussed at these initial planning meetings. Accuracy Assessment Points; • Create a Vegetation Field Key; Gradsect Meeting: • Write Local Descriptions for all A logical and systematic method of Vegetation Types; sampling the vegetation was necessary. The • Provide information and sections for Gradsect Sampling (GRADient-directed the Final Report. tranSECTs) approach was used and required the input of a number of different interested parties with a wide assortment of expertise to design the sampling method.

32 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Field Preparation Meetings: addition, we reviewed and discussed the Prior to beginning field work in 2002 and field data, aerial photo-interpretation, 2004, the project team met to discuss issues transfer to digital orthophoto quads for the approaching field seasons. These (DOQ's), and Park management needs. included developing an agenda for the orientation and training of the field crews, Accuracy Assessment Meeting: completion and application of the gradsect Following completion of the vegetation analysis, scheduling the plot work, and sampling, an accuracy assessment meeting defining appropriate Park contacts for was held on November 19, 2004 to discuss various issues that might arise. the accuracy assessment and the AA plot data collection. These included topics such Interim Status Meeting: as the methodology for defining the An interim project status meeting was held appropriate number and distribution (spatial following the completion of the 2002 and thematic) of the random plots, summer field season. The purpose of the development of the cost surface to be used meeting was for the project participants to in placing the plots on the map, data to be summarize their progress on the project and collected at the plot locations, and logistics to plan remaining tasks to be completed for completing the required number of plots prior to the start of the 2003 field season. in a single season. Topics covered during the meeting included progress on defining the Map Units, Preliminary Data Collection and comparison of the number of plots collected Review of Existing Information with their from the field and their distribution across the preliminary list of To minimize duplication of previous work associations, progress of checking the data and to aid in the overall mapping project, and entering it into the PLOTS database, existing maps and reports were obtained and the status of the fuels and photographic from various sources. The staff at ROMO data. We also briefly discussed schedules provided digital and hard copy background and tasks for the coming months. material for numerous themes including geology, fire history, soils and previous Map Unit Meeting: vegetation maps. Digital elevation models After a draft classification was developed, (DEM's), digital line graphics (DLG's), and representatives from the Park, BOR, CNHP, digital raster graphics (DRG's) were and NatureServe met at the BOR office in obtained from both ROMO and the USGS. Denver on February 7th 2003 to decide on The DEM's were manipulated to create appropriate map units. The purpose of the slope and aspect maps. meeting was to determine a final mapping classification for vegetation and land use at Before beginning the 2002 field season, ROMO. BOR presented a preliminary map CNHP reviewed existing literature and data unit schema after a review of the final from previous studies of vegetation in and vegetation classification and consideration near ROMO. Of particular interest was plot of prior work. Each map unit was discussed data presented in Peet (1981) which was and either lumped or split depending on the collected between 1972 and 1974, and a opinions of the various ecologists and GIS / large dataset collected by various remote sensing professionals present. In researchers from the Natural Resource

33 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Ecology Laboratory (NREL). Additional riparian vegetation classification project. plots presented by Carsey et al. (2003) were The 24 plots we used were originally also evaluated. The locations of all plots collected by Gwen Kittel in 1993 and 1996 evaluated from previous literature are from portions of the project area within shown in Figure 5. Larimer, Grand, and Boulder Counties. These plots were directly applicable to the The Peet (1981) data, although very project methods and were used in the pertinent, was collected with different classification as well as the mapping effort. assumptions and different methodologies. Instead of cover classes, Peet (1981) To establish a “preliminary list” of calculated stem densities for all trees and associations, NatureServe Ecologists shrubs inside the plot. Because stem counts queried the NatureServe BIOTICS Database do not equate with cover, it is impossible to for the types known to occur within the quantitatively use that data in the Southern Rocky Mountain Ecoregion. classification. In most instances we were Based on that query the preliminary list able to cross-walk the types described by contained 242 associations. This list was Peet (1981) to a recognized NVC type and then reviewed by a number of local experts assign the plots to a map unit. This allowed and refined to indicate the likelihood of a us to qualitatively use the plots to inform type occurring in the Park; 174 were the preliminary classification list and to considered probable to occur in the Park, 22 provide additional control points for the were considered possible to occur in the mapping effort. Park, and 46 were considered not probable to occur in the Park. Following further Several datasets were obtained from the discussion and revision at the scoping NREL plot database. These were similarly meetings we ended with a preliminary list evaluated for their usefulness, and were of NVC Associations likely to occur in applied to the classification and mapping in ROMO. Data from previous studies were a similar manner to the Peet (1981) data. also used to inform the development of this The NREL database included 180 plots that list, and as indicated above, to inform the were mostly collected using multi-scale final classification. At the start of the 2002 sampling methods (Stohlgren et al. 1996; field season, the preliminary classification Stohlgren et al. 1997; Stohlgren et al. 2000). included between 174-196 existing NVC association types. The plots from Carsey et al. (2003) were collected as part of a multi-year wetland and

34 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 5. Location of ancillary plots used for photointerpretative training and reference.

35 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Field Survey

Sampling Design: Stratified Random Gradsect Our ultimate goal at ROMO was to obtain a thorough description for the range of plant For ROMO’s modified gradsect, elevation, communities, both the common/extensive geology and solar insolation, were chosen and the rare/unique (Austin and Heyligers as the key abiotic factors (Table 6). We 1991). To this end we felt that an unbiased then split each gradsect variable into logical census of all the vegetation (i.e. a complete classes to best reflect the vegetation enumeration of the population) would not distribution and created digital map layers be achievable or practical for such a large, using ArcView GIS. These GIS layers were remote Park. As a result, to cost-effectively then added together to generate a map capture the full spectrum of vegetation we coverage of all combinations occurring in felt it necessary to optimally locate ROMO, with each unique combination sampling plots using “Gradsect Sampling” representing a Biophysical Unit (BPU). (GRADient-directed tranSECTs) (Gillison and Brewer 1985). Gradsects are a survey At ROMO there were 210 BPU types within method that addresses 1) the need for the model area that formed a mosaic of representative sampling based on 51,242 polygons. The model area extended environmental stratification, 2) the need for past the mapping boundary to the north, a compromise between statistical sampling, south and west and fell short to the east due practical logistical problems, and costs, and to limited geologic coverage. This model 3) the value of replicated and randomized area was subsequently clipped by the sampling (Austin and Heyligers 1991, mapping boundary and randomly sampled. Gillison and Brewer 1985). We assumed We selected a subset of these BPUs using a that a modified Gradsect methodology cost-surface analysis, (Appendix D) which would allow field crews to visit the full favored polygons that were more accessible, spectrum of physical environments and thus of adequate size, and spatially dispersed. most of the vegetation types. This resulted in 2-3 polygons of each type for a total of 925 polygons selected for For ROMO, we decided that a spatial- possible sampling during the initial field historical model coupled to a 30-meter season. At ROMO, polygons ranged from 1 digital elevation model (DEM) of the Park -10 ha in size, although the overall range would be more predictive of vegetative was 0.1-47 ha. This cost-surface process diversity and more efficient than a linear for selecting sampling locations was transect approach. A working group of especially important for ROMO, due to Park, CNHP, NatureServe and BOR access difficulties caused by the steep and ecologists / botanists / geographers familiar inaccessible areas of the Park. with the region selected the model’s driving variables; those thought to influence The crew response to field visits to the vegetation response. During this process, selected BPU’s was mixed to poor. Often practical constraints were also considered the comment from the crews was that the including the lack of time and money to BPU’s really did not provide anything new develop new digital data layers. or interesting and questioned the utility of

36 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park this approach. One positive note to the use of habitats. Often, once a crew reached the of BPU’s is that it provided us a utility to selected target they sampled other areas disperse the targets over a very wide range adjacent or near the modeled target.

Table 6. Environmental variables and classes used in the modified gradsect analysis for ROMO.

(For more detailed information on the ROMO National Park Analysis - Sample Site Selection Methodology see Appendix D.)

SOLAR ELEVATION (FT) INSOLATION GEOLOGY 6000 - 7000 10-Full Shade Granite of Hagues Peak 7000 - 8000 20-Partial Sun Troublesome Formation 8000 - 9000 30-Partial Shade Talus 9000 - 10000 40-Full Sun Alluvium 10000 - 11000 Granitic gneiss 11000 - 12000 Colluvium Till of Pinedale-age Silver Plume Granite Biotite Schist

General Plot Collection Considerations

At the beginning of the 2002 field season, The process of collecting vegetation plots is four crews of two people each were hired to long and complicated, and required several collect data in ROMO. The crews were months of planning and preparation. Based trained in the vegetation and fuels sampling on the results of the GRADSECT analysis, methodology and provided with BPU maps proposed plot locations were evaluated for identifying possible plot collection sites. distance from access points and difficulty of Crews were also given a list of the 180 travel. Plot locations in areas determined to potential vegetation types to be sampled and be more than three or four hours from an instructed to try to collect three plots in each access point were considered as type. Since ROMO is large, diverse, and backcountry, while those less than that were inaccessible by car in many places, crews considered as front country. All front were hired both for their botanical and country sites were accessed via day hikes ecological skills, and for their ability to work from the nearest road access point. effectively in variable outdoor conditions, Backcountry sites were accessed via foot and to work while backpacking. Crews were trail on multi-night backpacking trips. provided with training in project methods, Crews accessing backcountry sites camped housing protocols and vehicle use. in established campsites and cross-country camping zones. A total of 102 backcountry camp night reservations were required to

37 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

collect data from all of the backcountry sites. could choose to collect an Observation point Crews typically stayed in a single established or navigate to a different polygon. An site for 1 to 3 nights, and remained in the observation point is a short version of a plot backcountry for 3 to 7 nights. which uses a dimensionless plot to document the structure, dominant species, and The crews were provided with a field manual environmental attributes in the area. It can describing all of the methodology for the be collected using the same forms and plot sampling, as well as supplemental instructions as for full vegetation plots. information on backcountry safety, species Observation points are very useful for the lists, and accepted plant species codes. The photointerpreters as training / reference sites. field manual provided to the crews and a plot field form are provided in Appendix B. Along the way to and from the selected Examples of the field forms are in Appendix polygons, crew members would pay C. The following is a general description of attention to vegetation types they were the process. passing through. If they observed other needed vegetation types (especially rare Data Collection: Relevé Plots types) or found possible new vegetation Before leaving for the field each day, and types (undocumented vegetation before each multi-day trip, crews would plan composition which repeated on the a strategy for collecting plots most landscape) which might help inform the efficiently. They would take into classification of ROMO, they would stop consideration the proximity of selected and collect either a full vegetation plot, or an BPUs to roads, trails, and to each other, as observation point. If crews noticed weeds or well as topography and vegetation in the area rare plants, they would stop and record those to be surveyed, and would plan routes to as well. collect the most plots in the most different potential vegetation types without excessive At each sampling location, plot data were travel time. Crews would then gather all collected using the protocols of the NPS field equipment and personal gear needed for National Vegetation Mapping Program. At the duration of the trip. the plot center, crews would bury a permanent marker (a small copper tag The crews would navigate to the selected inscribed with the project acronym, plot area using GPS units as well as maps and code, and date, attached to a galvanized nail) compasses. After arriving at the selected and record the UTM location from the GPS. BPU, crews would select a specific location They would then lay out the plot, using to place the plot. They would walk through measuring tapes, according to the size specified in the field manual for that the polygon to get an idea of the vegetation. 2 If the polygon contained vegetation in a type, vegetation type (most plots were 400m ). or types that crews still needed to collect, Crews would begin analyzing vegetation by they would choose a relatively homogenous dividing the vegetation visually into strata, and representative area to place a plot. If the or height classes, and recording the dominant polygon was diverse, the crew might select species by cover in each stratum. They two or more locations in order to capture that would then develop a comprehensive species variation. If the polygon contained only list for the plot by recording the types that crews did not need to collect they

38 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park species name and percent cover for each collected included information on both live plant found within the plot. Numerous other and dead/down fuels. Live fuels data data describing the environmental included the density, DBH, height, crown characteristics of the site were collected at ratio, and height to the base of the crown for each plot including elevation, slope, aspect, all live trees in the stand, as well as the soil texture, surficial geology, percent cover of shrubs and herbaceous species. In ground cover, and hydrology. A complete particularly dense stands, crews were set of forest fuels data were also collected permitted to sub-sample a quarter of the full from each plot. Before breaking down the plot. plot, crews would attempt to place the vegetation into one of the potential For the dead/down fuels, crews recorded vegetation types. If the plot did not fit into a depth of litter and duff, shrub cover, vegetation type, they would assign a type herbaceous cover, cover of bare soil, and the based on the dominant species in the top cover of wood, litter, and duff. Four photos two strata. Two pictures showing the plot were taken at each site and photo center and typical vegetation were taken at information was recorded on the fuels each plot. When the Plot had been datasheet. Fire data collected during 2002 completed, crews would navigate to the next also included a surface fuels selected BPU and begin the process again. characterization, further information on dead and down wood and extensive Data Collection: Forest Fuels Data information, including DBH, on all trees Throughout both the 2002 and 2004 field found at the site (Field forms in Appendix seasons, fire specific data (fuels data) were C). collected at each site visited. Fuels data

39 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

40 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Aerial Photography Acquisition and Orthobase Map Development

A contract for aquiring aerial photography (1"=1,000') scale and 366 color photographs was let to Horizons, Incorporated of Rapid for the 1:40,000 (1" = 3,333') all printed on City, South Dakota. This contract specified 9"x 9" paper stock. Overlap for all the acquisition of color aerial photography photographs was approximately 50-60% for ROMO and vicinity at two different and sidelap between flight lines was scales. In addition to the standard 1:12,000 approximately 20-30%. All photography scale photography typically used for photo- was acquired between September 25 and interpretation, the project also required October 3, 2001. Repeat dates were 1:40,000 scale photography for the required due to the large area being production of orthophotos for the entire acquired in addition to the challenging area. terrain. Airborne GPS was collected with a base station at the Fort Collins, Colorado Normally for these types of projects Airport (Latitude: 40 26 58.07884; photography is acquired at mid-summer Longitude:-105 00 24.66076; however due to the emphasis on mapping Height:1509.710 m; Antenna ht:0.203 m). aspen the acquisition date was pushed back GPS points provided for a flight line point to September. The aspen leaf color change shape file and aided with the ortho- in the fall made photointerpretation of this rectification process. important map unit much easier. We did experience long shadows in some areas but The contracting for new digital and hard this was not a tremendous issue. In fact, in copy color ortho base maps for this project some cases the shadow and subsequent not only provided new base maps for the profile from the trees allowed us to Park but also provided for much better determine species. transfer of interpreted information to better base data. Technical specifications for the The project area is covered by 28 flight base map photography and subsequent data lines flown south to north for the 1:12,000 manipulation can be found in the orthophoto scale photography (Figure 6) and 10 flight metadata file located in the accompanying lines for the 1:40,000 scale photography DVD. All aerial and ortho-photography (Figure 7). Some of the flight lines overlap received a rigorous examination by BOR with repeat flights due to elevation and prior to being accepted. scale considerations. A total of 1,412 color photographs were taken at 1:12,000

41 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 6. Flight lines and photo centers for 1:12,000 color photography.

42 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 7. Flight lines and photo centers for 1:40,000 color photography.

43 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Photo Interpretation – Map Units

Photo Interpretation

Photointerpretation was done using the 9 x Stereoscopic photointerpretation 9, 1:12,000 scale color photographs in addition to the paper copies of the orthophotos. The interpretation on the aerial photography or the paper copies of the orthophotos depended a lot on the complexity of the landscape. Areas with high heterogeneity were mapped using the aerial photos and the orthophotos as ancillary. Areas of high homogeneity, such as vast areas of young lodgepole, were mapped with less reference to the aerial photography. Map Units and Polygon We requested that the orthophotos be Attribution delivered from the contractor (Horizons Inc) as 48, 1:12,000 scale sheets and The map units delineated on the printed on photographic paper with a 1,000 orthophotos were derived from the NVC meter UTM grid. These were then covered classification as constrained by the with translucent (semi-frosted) Mylar, limitations of the photography. After an fastened together, and backlit on a light initial map unit meeting with all parties to table. All UTM grid points were marked this mapping effort, we arrived at an initial on the overlays and the initial polygons list of map units we thought reasonable. In were delineated using a 0.5 mm lead pencil. few instances, one NVC association This served as our base for future transfer corresponded to one map unit. In most to the digital realm. cases, however, many associations were grouped into one map unit. The number Mylar overlays placed on each aerial photo and names of map units changed as the allowed us to make notes and delineate project progressed. (see Table 12 in polygons. At this stage of interpretation we “Results” section) After the initial used a stereoscope to help recognize determination of preliminary map units the complex photo signatures and three- photointerpreters started their work. After dimensional features on the 9X9 aerial some PI had been done several additional photos. We then manually transferred these map units were added as the to the orthophotos (Figure 8). Finally, in photointerpreters became more familiar order to insure completeness and accuracy, with the project. Some new map units were digital transfer specialists reviewed all of an attempt to use mental models rather than the interpreted orthophotos for consistency just signature as a photointerpretive tool. and recommended changes where After the accuracy assesment the number of necessary. Once all the obvious vegetation map units was again modified. A map unit and land-use classes were delineated we key is included in Appendix G. proceeded into the second stage.

44 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

The photointerpretation included (Jenness 2005). A grid was used for the vegetation context and structure. Four slope figure in this dataset. attributes are included. These are map unit, height, density and coverage pattern. Acres and hectares are calculated using The structural categories and codes are XTools Pro for ArcGis Desktop listed in Table 7. A lookup table for the (www.xtoolspro.com). map unit names is included in the attached dataset and in Appendix J. There is an Instrumental to the photointerpretive effort oldvegcode item that reflects the earlier was the use of the GPS located vegetation map unit classification. plots collected by the field crew. These plots gave us a good idea of what the Each polygon has a number of attributes signatures of the individual map units that are stored in the associated table should look like. In addition to the within the GIS database. Many of these tablular data associated with each attributes are derived from the vegetation plot were the four photographs photointerpretation and others are collected at each plot. These photographs calculated or crosswalked from other helped not only in identifying the classifications. Table 8 shows all the immediate area but also provided us with a attributes and their sources. “look” at the areas surrounding the vegetation plot which might be a different Anderson Level 1 and 2 codes are also map unit. In addition to the vegetation included. These codes should allow for a plots we made use of a number of different more regional perspective on the sources that had ground referenced vegetation types. A lookup table for the information. names associated with the codes is included with the attached dataset and in Additional ground information was Appendix J. supplied by a wide range of researchers that have done previous vegetation work There is also an ecological system code in the Park. Many plots were provided by (El_Code). Look-up tables for these Dr. Tom Stohlgren and the NREL at ecological system names are also included Colorado State University in Fort Collins. with the attached dataset and in Appendix Additional plot data were provided by the J. There are three El_codes in the Fire program at ROMO. Others included database. These reflect the one to many Colorado State Parks and Gwen Kittel relationships that may exist when cross (Natureserve). The largest share of walking the Map Unit to ecological previous plot data were provided by Dr. systems. Robert Peet who collected 305 plots for his dissertation (Peet 1981). The Slope (degrees), aspect and elevation are distribution of the ancillary plots tended calculated for each polygon label point toward that area around the principal roads using a DEM and an ArcView script (Figure 5). The content of each of the developed by Jenness enterprises and plot data sets varied from simple downloadable from www.jennessent.com. references to the overstory to detailed The slope figure will vary if one uses a species list and cover values for each TIN versus a GRID for the calculation species. These data are not included with the ROMO dataset as the data were loaned

45 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park to this project to help with the wishing to use the additional data are photointerpretation and the specifics encouraged to contact the individual remain the intellectual property rights of researcher directly. the individual researchers. Those readers

Table 7. Structural categories for vegetation photointerpretation.

Code HEIGHT 1 < 1 Meter 2 1 - 5 Meters 3 5 - 15 Meters 4 15 - 30 Meters 5 > 30 Meters COVERAGE DENSITY 1 Closed Canopy / Continuous 75 – 100 % 2 Discontinuous 50 – 75 % 3 Dispersed 25 – 50 % 4 Sparse < 25 % COVERAGE PATTERNS 1 Evenly Dispersed 2 Clumped / Bunched 3 Gradational / Transitional 4 Alternating

Figure 8. Example of transfer process from aerial 9 x 9 photography to color orthophoto base map. (Example from – Vegetation Mapping Project (Cogan et al. 2003).

46 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 8. Polygon attribute items and descriptions used in the ROMO spatial database (GIS coverage).

Attribute Description AREA* Surface area of the polygon in meters squared PERIMETER* Perimeter of the polygon in meters ROMO_VEG#* Unique internal polygon coding ROMO_VEG-ID* Unique internal polygon coding VEGCODE Final Map Unit Codes - BOR derived, project specific. OLDVEGCODE Initial Map Unit Codes - BOR derived, project specific. HEIGHT Height range of the dominant vegetation layer. (Height classes: 0-2, 2-5, 5-10, 10-20, >20 meters) DENSITY Density of the tallest strata. (Density classes:<25%, 25-50%, 50-75%, >75%) PATTERN Vegetation pattern within the polygon. (Vegetation pattern classes: Evenly dispersed, Clumped/bunched, Gradational, Alternating) SLOPE Slope of label point within polygon (degrees). ASPECT Aspect of label point within polygon. AND_LEV1 Land Use and Land Cover Classification System (USGS, Anderson et al. 1976) Level 1. AND_LEV11 Land Use and Land Cover Classification System (USGS, Anderson et al. 1976) Level 2. HECTARES Area in Hectares ACRES Area in Acres ELEV_M Elevation in meters for label point ELEV_FT Elevation in feet for label point EL_CODE_1 Ecological Systems Classification Code - NatureServe Ecological Classification. EL_CODE_2 Ecological Systems Classification Code - NatureServe Ecological Classification. EL_CODE_3 Ecological Systems Classification Code - NatureServe Ecological Classification. (*ArcInfo© default items)

47 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Digital Transfer

The transfer process for ROMO involved points (the UTM grid) as marked on the taking the interpreted line work and orthophotos to the same ones in the digital rendering it into a comprehensive digital images. In this manner the transfer was 1- network of attributed polygons. To to-1. accomplish this, we created an ArcInfo© GIS database using in-house protocols. Once scanned and registered, we removed The protocols consist of a shell (master all erroneous information such as dangling file) of Arc Macro Language (AML) lines. After cleaning, we joined the lines scripts and menus (nearly 100 files) that into polygons by building topology in the automate the transfer process, thus insuring GIS program. The resulting polygons that all spatial and attribute data are were then edge matched with those from consistent and stored properly. The actual adjacent orthophotos. Finally, we created transfer of information from the interpreted labels for each polygon and use these to orthophotos to a digital, geo-referenced add the attribute information. Using this format involved scanning, rasterizing, process we created one final coverage or vectorizing, cleaning, building topology, spatial database for the entire project. and labeling each polygon. Attribution for all the polygons at ROMO included information pertaining to map The scanning technique involved a multi- units, NVC associations, Anderson land- step process whereby the Mylar overlay use classes, and other relevant data. sheets produced by the photo interpreters were scanned into a digital form. The Attribute items requested by the ROMO digital image file (tagged image format .tif) fire program included height, density, and created from the scanned sheet was then pattern. All of the attribute items are listed converted from a raster image to a vector in Table 8 and are referenced in the file using RSGIS-developed AMLs in ROMO vegetation look-up table included ArcInfo©. The vector file was then geo- on the accompanying CD-ROM. Attribute referenced to the matching digital version data were taken directly from the of the orthophoto. The essential principle interpreted photos or were added later of geo-referencing was to match control using the orthophotos as a guide.

Scanning a ROMOorthophoto sheet

48 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Plot Data Management and Classification Analysis

Plot Data Management data were statistically analyzed using direct Following the field season and prior to data and indirect ordination and a polythetic entry, all plot forms were checked to ensure divisive technique and assigned to quality control (QC). Particular attention vegetation types. The statistical analysis was paid to making sure that the recorded was conducted with the aid of the multi- plot location was correct and that all variate statistical software package PC- relevant fields were filled in. When ORD version 4.14. An additional 9 plots information was missing, an effort was and 8 observation points were collected in made to find and record that information, the fall of 2004 to augment sampling of often from the associated fuels form, or targeted vegetation types, but were not used from other data sheets produced by the in the multi-variate analysis. same crew on that or an adjacent day. Changes to field form entries were made in After subjecting plot data to QC, we began red pen and marked with a date and the the process of analyzing and interpreting the reviewer’s initials. data. The goal was to classify the plots at the community level based on species Following the QC of the datasheets, the data composition and environmental were entered into the PLOTS database, and characteristics. Our intentions were to all plots were subjected to a second QC to finalize the community classification by eliminate any data entry errors. During this merging our ecological understanding of the second QC, the database was examined, classification system with two multivariate sorted, and queried to find missing data, analysis techniques, Two-way Indicator misspellings, duplicate entries, and typos. Species Analysis, or TWINSPAN (Hill The species lists were carefully examined to 1979), and Detrended Correspondence make sure that only USDA PLANTS Analysis, or DCA (Hill and Gauch 1980). (USDA, NRCS 2005) names and acronyms were used, and that species names and A data matrix was created in PC-ORD using assignments to strata were consistent and data from all 547 plots. We first ran the logical. Plant lists were compared to the matrix in TWINSPAN, in the hopes that we assigned association name to assure would see some initial breaks, which would correlation. allow us to further refine our analysis using DCA. We hypothesized that the initial Vegetation Classification breaks would be based on high coverage indicator species such as Abies lasiocarpa, Field crews collected data from 632 sample Pinus contorta, Pinus ponderosa, Salix plots during the summer of 2002 (Figure 9). planifolia, Caltha leptosepala , Carex Of the 632 plots collected, 547 were full aquatilis, and others. The initial analysis vegetation plots and 85 were the shorter reflected this trend; however, there were observation points (see Appendix C for field other, less indicative species, such as forms). Data from these plots were Achillea millefolium and Taraxacum reviewed for accuracy and entered into the officinale, which occurred with such PLOTS database during the fall of 2002. regularity that unnecessary breaks were From November 2002 to January 2003, the created by the analysis.

49 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

After the initial run with TWINSPAN we similar community types, and make a created a priori breaks in our data set based decision based on this information and on on the physiognomic classes of the plots. our ecological understanding of these We hoped to decrease the confusion and communities. The DCA was helpful as a lack of consistency inherent in an analysis tool to bring out patterns we would of such a large number of plots and plant otherwise not discern from the data; species. We created three subsets of data: however, the final call for classification was Herbaceous plots, Shrubland plots, and based primarily on information regarding Forest plots, and ran each subset as a unique previously studied vegetation community data set on DCA with default settings. types and on our ecological interpretation of the plot data. Data were interpreted through graphical representation on two axis. Each iteration After all plots had been classified to NVC would display “outlier” plots, or plots which vegetation types, local descriptions were visibly stood apart from the larger written for each type, and a dichotomous congregations of plots on the two way axis. key to the vegetation types of ROMO was We would manually analyze the data behind written (Appendix E). The local outlier plots on the graph, and classify the descriptions are based on the plot data from plots or table them for further discussion, ROMO only and describe the structure, then remove them from the analysis. The composition, and environmental judgment of what was and was not an characteristics for the type as it occurs at outlier was up to the interpretation of the ROMO. Because the descriptions are based analyzer. After removing outliers, and, in only on the ROMO field data, their effect, diminishing noise in the matrix, we completeness and accuracy is solely a would run the analysis again. This iterative function of the number of plots and process quickly revealed groupings of plots. observation points collected for each At this point we would manually analyze particular type. Additional plots in any the data behind the plots in the groupings. given type can further inform the During analysis we took into account classification of the type and its description. environmental characteristics of the plot and species composition. We also looked at the The field key combines the characteristics provisional community name given to the indicated by the ROMO plot data with the plot by the field researchers, as this was a essence of the NVC concept for each of the helpful starting point for further analysis, associations. It provides the user with a and a useful piece of information for series of dichotomous choices that result in making final tough calls. identification of the association. The field key was used during the 2004 Accuracy The limitations of DCA were apparent, as Assessment to determine the vegetation the groupings often did not easily reflect the type for each of the AA Points. Based on final classification of the plots. When the results of its use during AA, the key was possible we would reference previously edited to include previously omitted types published materials regarding relevant and and to clarify the text and simplify its use.

50 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Map Verification and Accuracy Assessment

As we completed the ortho-photo interpretation and digital transfer for sections of the Park, draft 1:12,000-scale Sample Method hard copy vegetation maps were printed for review. In all cases, we checked The accuracy assessment protocol takes these draft maps against the interpreted into consideration maximum and photographs to ensure that the polygons minimum sample sizes. Considerations were labeled properly and to locate any include statistical as well as cost extra or missing lines. We also constraints and mapped class abundance compared the map labels to the plot data and frequency. The sample selection is a if they fell in the same location. Copies stratified random sample, stratified by of the revised draft map were then sent map units. Five scenarios are based on into the field on several occasions by the class abundance and frequency and are photo interpreters for ground-truthing. defined in Table 9. During the ground-truthing process, we verified aerial photograph signatures Sample Site Selection using landmarks and GPS waypoints. The map and map units were then These parameters were coded into in- modified to correct any mistakes. house software programs that allows for repeat sample selection using a variety Sample Design of sample choices such as cost weighting and distance from polygon boundary. Selection of AA sample points followed The cost weighting allowed us to that described by the NBS/NPS eliminate sample points that had Vegetation Mapping Program, Accuracy extremely arduous access Assessment Procedures manual. The (distance/difficulty = cost) or were in design attempts to adhere to scientific dangerous locations. Being able to principles that govern sampling and choose minimum distance to polygon statistical analysis and also be practical. boundaries helped to eliminate ecotonal The consideration of map accuracy boundaries which lead to confusion and typically can have two components: loss of effort. A minimum distance of thematic map accuracy and positional 10 meters was chosen for this effort. accuracy. The accuracy assessment that The distribution of sample points is follows reflects only thematic map shown in Figure 10. accuracy. Positional accuracy is not considered. Given that polygon AA Sample points were collected for the boundaries are only occasionally “hard” entire mapping area as opposed to the and subject to interpretation it makes original vegetation sampling which was little sense to spend the effort to quantify restricted solely to the Park itself. a subjective boundary. Field crews were provided with two sets of samples. The primary set included the preferred target for the sample selection. If a target was inaccessible for

51 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park any reason, the crews were free to used as AA Points. The field crews substitute from a secondary set of points. were instructed to navigate to these The effect of this arbitrary reselection points and complete an AA datasheet reduces somewhat the stratified random (Appendix C). Given the time frame of selection of points. The positive effect the project, and the rugged nature of the of this is to take advantage of the cost of Park, it was assumed that not all of the sending a crew to a particular location. generated points would be accessible. A secondary set of 1,010 possible AA Data Collection – AA Points Points was included on the maps as replacements for those primary points Field maps were produced that showed which might be discovered to be the sample point and polygon boundary. inaccessible or otherwise unusable. The addition of the polygon boundary to the field map aided in navigation to the Between June 5th and October 15th point and provided the field crews with 2004, the Field Crews collected 1219 some contextual information. Field AA Points. 811 AA Points were from crews navigated to each point using the the primary point list and 408 were from field maps produced for this effort in the secondary list. Figure 10 shows the addition to a GPS with a known target locations of the collected AA points. location. Each day crews choose points to visit based on logistical factors. Low To help control cost and logistic issues elevation points were visited early or late only those map units that had a in the season, while high elevation sites vegetative component received an were only accessible when not covered accuracy assessment. This resulted in by snow. Remote sites were visited the removal of 5422 polygons from during backpacking trips, which were consideration or 15% of the total number planned prior to the start of the field of polygons. This translates into 11% of season. Points located around Lake the mapped area not accuracy assessed. Granby were accessed by pontoon boat, Un-assessed map units include areas while all other sites were accessed by car such as rock outcrops, glaciers, water and/or on foot. Field days were planned bodies etc. These areas were assumed to around collecting as many primary have a very high accuracy, as there is points as possible; however, when little, if any, ambiguity in their secondary points occurred along a interpretation. The result of this is that planned route for the day, they were the reported overall accuracies are most surveyed in anticipation of future points probably lower than the actual overall which might be missed. The tally of map accuracy. which points had been collected in each Map Unit was updated throughout the In June of 2004, the CNHP Accuracy summer. During the last few weeks of Assessment field crews were given the project, areas for the crews to visit printouts of the aerial photographs and were chosen strategically, to assure point topographic maps overlaid with the map coverage across all of the Map Units. unit polygons. These maps contained 1167 randomly selected locations to be Upon arrival at a point, crews would begin with a broad visual survey of the

52 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

area. This was done to determine subjected to another round of QC to whether vegetation at the point was catch data entry errors. The Map Unit representative of the Map Unit polygon was not specifically assigned in the field, (ecotone or inclusions). If vegetation but was assumed to correlate directly was not representative, the crew would with Association; with one Map Unit move the point to a more representative encompassing one to several location within the polygon and record Associations. A Map Unit column was the distance and bearing to the new added to the database and filled using a point. The crew would then visually Microsoft Access query. Those points determine the boundaries of the point to which did not fit well into an existing be sampled. The minimum mapping NVC Association were keyed to Map unit is ½ Ha and this was used as the Unit by hand. sample plot. Crews would then begin collecting data on species composition, vegetation structure, and geology and topography of the area. After filling out the AA Point form, the crew would use the Field Key (Appendix E) to assign an NVC Association to the plot. If no Association seemed to fit, the crew would assign an association name to the plot based on the NVC naming conventions for Associations (dominant species of the primary strata). At each plot four pictures were taken in each of the cardinal directions from the plot center. Crews were instructed to document what they observed at the plots by recording extensive field notes. The pictures and the notes that crews collected in the field proved very useful in resolving classification questions later during the AA.

At the end of the field season, all AA point paperwork was subjected to the same quality control (QC) procedures as the vegetation plot data. While all fields on the AA form were checked for accuracy, particular attention was given to checking the UTM’s and plot numbers, and to comparing the assigned association name with species data. All AA point data were then entered into the PLOTS database. Following the data entry, the AA data in the database was

53 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 9. Location of vegetation plots collected at ROMO.

54 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 9. Recommended map accuracy sample number per class by frequency and area.

Area Recommended Polygons Scenario Description occupied number of in class by class samples in class Scenario A: The class is abundant. It covers more than 50 hectares of the total area and consists of at least 30 polygons. In this case, the recommended sample size is 30. >30 > 50 ha 30 Scenario B: The class is relatively abundant. It covers more than 50 hectares of the total area but consists of fewer than 30 polygons. In this case, the recommended sample size is 20. < 30 > 50 ha 20 The rationale for reducing the sample size for this type of class is that sample sites are more difficult to find because of the lower frequency of the class. Scenario C: The class is relatively rare. It covers less than 50 hectares of the total area but consists of more than 30 polygons. In this case, the recommended sample size is 20. The rationale for reducing the sample size is that the class occupies a small area. At the < 30 < 50 ha 20 same time, however, the class consists of a considerable number of distinct polygons that are possibly widely distributed. The number of samples therefore remains relatively high because of the high frequency of the class. Scenario D: The class is rare. It has more than 5 but fewer than 30 polygons and covers less than 50 hectares of the area. In this case, the recommended number of samples is 5. The rationale for reducing the sample size is that the class consists of small polygons and the frequency of the polygons is low. Specifying more than 5 sample sites will 5, 30 < 50 ha 5 therefore probably result in multiple sample sites within the same (small) polygon. Collecting 5 sample sites will allow an accuracy estimate to be computed, although it will not be very precise. Scenario E: The class is very rare. It has fewer than 5 polygons and occupies less than 50 hectares of the total area. In this case, it is recommended that the existence of the class be confirmed by a visit to each sample site. The rationale for the recommendation is that Visit all and < 5 < 50 ha with fewer than 5 sample sites (assuming 1 site per polygon), no estimate of level of confirm confidence can be established for the sample (the existence of the class can only be confirmed through field checking).

55 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 10. Distribution of accuracy assessment points, ROMO and environs.

56 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

AA Metrics

Once all the AA data had been entered number of “correct” observations and compiled the accuracy analysis divided by the sum of the row. portion of the project was started. This involved a number of steps including an Overall accuracy for the map was initial binary accuracy assessment, calculated as: calculation of confidence intervals, a k fuzzy evaluation of the AA data, n hypotheses testing and the construction ∑ ii of fuzzy error distribution map. =1i n

Binary accuracy assessment: All AA where k is the number of land cover plots and their respective map unit types and n is the total number of classification (reference layer) were reference points. This formula is simply compared to the digital vegetation the sum of the diagonal entries divided polygon data (predictive layer). This by the total number of AA points. provides an initial overall accuracy assessment and omission and Confidence Interval: The 90% commission errors (User’s and confidence interval for a binomial producer’s accuracy respectively). distribution is obtained from the (Unless otherwise noted all subsequent following equation: formulas are described from Accuracy Assessment Procedures, 1994) ⎪⎧ ˆ ()1p − ˆp 1 ⎪⎫ ˆ ±⎨zp α + ⎬ User’s accuracy is calculated as: ⎩⎪ n ()n2 ⎭⎪

n where za = 1.645 (this comes from a ii n table of the z-distribution at the i+ significance level for a two-sided limit

with a 90% confidence interval), p^ is the where i is the land cover type, nii is the number of matches between map and sample accuracy (0 to 1.0) and n is the number of sites sampled. The term reference data and ni+ is the total number of samples of i in the map. This formula 1/(2n) is the correction for continuity. is the number of “correct” observations The correction should be applied to divided by the sum of the column. account for the fact the binomial distribution describes discrete Producer’s accuracy was calculated as: populations.

n A kappa statistic is calculated for overall ii accuracy for each fuzzy level evaluated n i+ as follows:

where n+i = total number of sample of i Kappa can be used as a measure of in the reference data. This formula is the agreement between model predictions

57 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

and reality (Congalton1991) or to each of the AA plots and assign a fuzzy determine if the values contained in an level to that plot. Fuzzy level error matrix represent a result designations are shown in Table 10. The significantly better than random (Jensen result of this fuzzy designation then 1996). Kappa is computed as allowed us to evaluate each fuzzy level r r using the standard binary approach. That is, we developed a contingency Ν xii − ++x xx +ii ∑∑() table for fuzzy levels 5, 4 and 3. k = ==1i 1i Because we are only interested in the 2 x fuzzy levels that allow for varying Ν −∑()xx ++ ii degrees of membership and still be considered correct we ignored fuzzy where N is the total number of sites in levels 2 and 1. the matrix, r is the number of rows in the matrix, xii is the number in row i and To facilitate this analysis it was column i, x+i is the total for row i, and necessary to create one table that xi+ is the total for column I included all the information necessary (Jensen1996). Existing Arcview scripts for a multiple contingency table analysis. made this onerous process easy and A screen shot of this table is shown in repeatable (kappa_stats.avx by Jenness Figure 11. The table has several and Wynne (2004) or kappa.avx columns. The first column is generated developed by the RS/GIS Laboratories at by the GIS software and may be ignored State University (2003) and for analysis purposes. The second available at column is the unique plot id for each AA http://www.gis.usu.edu/~chrisg/avext/. point. Column three is the map unit assigned to the AA point. After each Fuzzy Accuracy Assessment: The AA point was collected it was assigned need for an alternative to the standard to a vegetation association and binary approach of accuracy assessment subsequently to a map unit. The was recognized some time ago. Gopal number in column three is the equivalent and Woodcock (1994) described the first of the map unit code. This column fuzzy accuracy assessment approach that becomes the “reference” point when is commonly used today. This type of applied to a binary analysis such as analysis allows for degrees of described in the binary approach above. membership to a particular class. That Columns four and five are the UTM is, we are allowed to recognize that a coordinates for each AA point. Column particular class may be considered six is the fuzzy designation applied to wrong using a strict binary approach but each plot. The designation of a fuzzy with the fuzzy analysis that class may be code is a lengthy process and made more mostly correct. This does provide a difficult by its subjective nature. To much better representation of the provide some objectivity to the fuzzy continuity present in the real world and membership we developed some on the still allows us to map using discrete fly rules for evaluating the plots and the classes. subsequent fuzzy designation. The rules are rough in that they do not evaluate The standard approach to assigning every single possible permutation of fuzzy membership to a class is to review

58 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park species occurrence, density, canopy etc. and fuzzy membership. Consensus was An example of this process is as follows: reached on each designation before moving on to the next AA plot. If an AA plot is designated as map unit 7 (alpine meadow) and the vegetation map All plots that received a “correct” shows map unit 2 (alpine fellfield) then designation during the initial binary the initial binary accuracy assessment assessment received a fuzzy membership would indicate that this was wrong. 5. All other plots were then selected for During the fuzzy accuracy assessment other fuzzy membership designations. and redesignation we applied the following rules: To evaluate fuzzy levels 4 and 3 using a contingency table we recalculated the If there was no herbaceous or rock in the reference layer (AA Plot) to be equal to plot then it received a fuzz code of 1 – the predicted layer (digital polygon map) Absolutely Wrong: This answer is depending upon the fuzzy level. To do absolutely unacceptable. Very Wrong. this we created three additional columns If the cover was herbaceous but in the AA table. These are columns contained no cushion plants or rock “fuzz5, fuzz4 and fuzz3”. Evaluation of cover then it received a fuzz code of 2 – fuzzy level 5 is identical to the initial Understandable but Wrong: Not a good binary accuracy assessment as described answer. There is something about the above. The designation of values in the site that makes the answer fuzz4 and fuzz3 columns is as follows. understandable but there is clearly a If a plot received a fuzzy designation of better answer. This answer would pose 4 then all levels at and below 4 were a problem for the users of the map. If calculated to equal the polygon code. the plot had an herbaceous cover, This column was then used as the perhaps some cushion plants and had a “reference” layer in the contingency high amount of surface rock then it table. These would show up as “correct” received a fuzz code of 3 – Reasonable in the contingency table for fuzzy level or Acceptable Answer: Maybe not the 4. However, at fuzz5 this would still best possible answer but it is acceptable; show up as wrong which is what we this answer does not pose a problem to would expect for the very stringent class the user if it is seen on the map. membership at fuzzy level 5. When we Correct. If there was a high cover of calculate the contingency table for fuzzy cushion plants but not enough to be level 3 this AA plot would show up as called a fellfield as per the association correct because it received a fuzzy code description it received a fuzz code of 4 - of 4 and therefore it follows logically Good Answer: Would be happy to find that it must also be correct at fuzzy level this answer given on the map. Very 3. Adjacent to the fuzz5, fuzz4 and Right. If the AA point (reference point) fuzz3 columns is the “vegcode” which is matched the vegetation polygon the code for the predicted value (polygon (predicted value) then the fuzz code was layer). Also in this table is a column for calculated to 5 - Absolutely Right: No the original AA code assigned to each doubt about the match. Perfect. AA plot (“AAMU” – accuracy assessment Map Unit). This column is During the evaluation process, all three identical to the fuzz5 column and is authors debated the merits of each plot duplicated for the sake of clarity. The

59 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

“fuzzcode” column is the fuzzy contingency tables are presented in the designation for each plot arrived at by “Results” section of this report. plot evaluation by the authors. The

Table 10. Fuzzy set accuracy ranks (Gopal and Woodcock, 1994).

Fuzzy Description Class 1 Absolutely Wrong: This answer is absolutely unacceptable. Very Wrong 2 Understandable but Wrong: Not a good answer. There is something about the site that makes the answer understandable but there is clearly a better answer. This answer would pose a problem for the users of the map 3 Reasonable or Acceptable Answer: Maybe not the best possible answer but it is acceptable; this answer does not pose a problem to the user if it is seen on the map. Correct 4 Good Answer: Would be happy to find this answer given on the map. Very Right 5 Absolutely Right: No doubt about the match. Perfect

60 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 11. Example of fuzzy re-designation for the creation of contingency tables.

Hypothesis Testing In order to accept or reject this hypothesis we use the confidence The purpose of the hypothesis test for interval. There is an extremely close this accuracy assessment is to determine relationship between confidence weather or not the accuracy estimate intervals and hypothesis testing. When a exceeds 80% (program standard). For 90% confidence interval is constructed, the purposes of this accuracy assessment all values in the interval are considered we use the following hypotheses: plausible values for the parameter being estimated. Values outside the interval are “The hypothesis that 80% accuracy has rejected as implausible. If the value of been met will be accepted unless the the parameter specified by the null sample map accuracy is low enough so hypothesis is contained in the 90% that the conclusion that rejection is interval then the null hypothesis cannot appropriate can be drawn with some be rejected at the .01 level. If the value predetermined degree of certainty.” specified by the null hypothesis is not in the interval then the null hypothesis can be rejected at the .01 level.

61 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Error Distribution unsampled locations using randomly selected locations as reference points. In The distribution of error over the this case, the randomly selected points landscape was produced using ordinary are the 1207 accuracy assessment points kriging and the following parameters: collected for this project. Key to this analysis is the assumption that there Selected Method: Ordinary Kriging exists spatial dependence in the variable Output: Prediction Map to be predicted (Thomas and Jacobs 2000). The following analysis follows Number of datasets: 1 that described by Thomas and Jacobs (2000). Number of Points: 1207 The output figure is described in the Semivariogram/Covariance: “Results” section, Figure 13. Model: 0.21393*Spherical(5833.3)+1.2296 * “The fuzzy set reference data must be Nugget explored and modeled prior to kriging. Error modeling: The spatial autocorrelation of the fuzzy Microstructure: 1.2296 (100%) set reference data must be calculated Measurement error: 0 (0%) (semivariance) and plotted to understand the extent to which neighboring Searching Neighborhood: accuracy ranks influence each other. Neighbors to Include: 5 or at least 2 Positive spatial autocorrelation indicates for each angular sector that similar values are clustered together Searching Ellipse: in space, and is usually evident in Angle: 0 geostatistical data, or natural processes Major Semiaxis: 5833.3 (Odland 1988).” (Thomas and Jacobs Minor Semiaxis: 5833.3 2000). The semivariogram and spherical Angular Sectors: 4 model is shown in Figure 12. The yellow line in the semivariogram The usefulness of this tool is to be able represents the model to infer the accuracy of a map at

62 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 12. Semivariogram and spherical model for 1207 AA points in ROMO.

63 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

RESULTS type only to the Alliance level of classification.

Field Data Collection Using the methods described above, the vegetation plot data collected in 2002 Field data were collected during the were classified into 172 distinct summers of 2002 and 2004. During the vegetation types based on species summer of 2002, a total of 635 composition, structure, and vegetation plots were established in the environmental characteristics. Of these, Park. Of these, 547 were permanent 167 are recognized NVC types, while an vegetation plots and 85 were observation additional 5 are “Local” types specific to points. During the summer of 2004, a the Park, but not yet recognized in the total of 1219 accuracy assessment points NVC. The 172 classified types included were established. These data were used 120 of the types that were on the original to develop the classification of the preliminary list and 47 types that were vegetation of ROMO as well as verify not. Of the 120 types on the original list, the accuracy of the completed map. The 101 were originally considered accuracy assessment points also served probable, 11 were considered possible, to refine the classification by verifying and 8 were originally considered not the presence of additional types that had probable. Of the 47 classified types that not been identified during the vegetation were not on the preliminary list, 28 were plot sampling in 2002. An additional 9 existing NVC Associations, while 19 are plots and 8 observation points were new or Local Associations. collected in the fall of 2004 to augment sampling of targeted vegetation types, Assignment of all the accuracy but were not used in the multi-variate assessment sampling points to either a analysis. classified NVC type, or an “other/unclassified” category identified Vegetation Classification an additional 35 types (22 Associations, 7 Alliances, and 6 non-vegetated types). The preliminary classification produced These “Other” types have not been in the spring of 2002 prior to any field included in the classification, because sampling included 174 -196 vegetation they are based only on AA sample points types depending on the probability of and were not used to create the occurrence (probable, possible, or not classification. With additional sampling probable). These were types existing in and classification work these additional the NVC at the time and for which local types may represent possible future experts had reasonable certainty would extensions of the classification, or may occur in the Park. The analysis of the be recognized as variants of existing vegetation plots (and accuracy classified types. The list of these other assessment points) identified some of types is included in Appendix L. The those types as well as others not on the Dichotomous Key to the Vegetation preliminary list, but failed to identify Types of ROMO is located in Appendix others that were thought to be probable E. Local and global descriptions for to occur there. Additionally, some of the each NVC type can be found in data collected was sufficient to identify a Appendix I.

64 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

The vegetation described from Rocky and sub-alpine meadows, on wetter sites Mountain NP is very diverse and dominated by species adapted to classified types span all of the saturated and flooded soil conditions, or elevational zones and environmental as alpine tundra. The most diverse gradients found in the project area. They group of herbaceous communities is range from the foothills zone shrublands those that occur on wet sites and which and woodlands found on the east side of are dominated by sedges (Carex spp.) the park, though the montane and sub- and rushes (Juncus spp.) alpine forests of the east and west sides of the divide, to the alpine tundra and Shrubland types identified in the Park cushion plant communities found at the range from xeric foothill communities of Park’s highest elevations. They include bitterbrush (Purshia tridentata) and upland, wetland, and aquatic sagebrush (Artemisia tridentata Nutt. communities as well. ssp. Vaseyana) to carrs dominated by various species of willow (Salix spp.), to Forests and woodlands comprise the sub-alpine and alpine dwarf shrublands greatest number of classified types found dominated by huckleberry (Vaccinium within the Park. Together, these two spp.) or arctic willow (Salix arctica). In related categories account for 68 of the total, shrublands account for 49 of the 172 classified types. The classification 172 classified types. The most diverse presented here includes 48 forest types and prevalent group of shrublands are and 20 woodland types. On an aerial those dominated by species of willows basis, forest and woodland types also (Salix spp.) which total 27 associations. occupy the greatest spatial extent of all types described in the project area. Of Photographic database these, 19 types are dominated by spruce and fir (Picea engelmannii and Abies The photographic dataset collected in lasiocarpa), 13 are dominated by documenting the vegetation plots in ponderosa pine (Pinus ponderosa), 10 2002 and the AA points in 2004 includes types are dominated by lodgepole pine a total of 7,979 digital photographs. The (Pinus contorta), 5 are dominated by photos collected in 2002 at the Douglas fir (Pseudotsuga menziesii), 2 vegetation plot sites typically include six of the types are dominated by limber photos of each plot and total 3,080 pine (Pinus flexilis), and 20 are different frames. Four were taken on the represented by aspen and cottonwood cardinal directions from the plot center (Populus teremuloides, P. balsamifera, and two additional representative photos and P. angustifolia). Taken together, were taken with bearings recorded. The these associations represent the largest photos taken in the 2004 season typically portion of the Park’s vegetation types. include 4 photos of each plot and total 4,899 different frames. These photos A total of 49 herbaceous community were taken on the cardinal directions types were classified for the Park. only. Herbaceous communities include graminoid dominated as well as forb The dataset of digital photos has been dominated types, and may occur as xeric provided to the Park GIS Manager for to mesic foothills types, mesic montane

65 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park integration into a database and map that will allow users to easily query and display the photos.

66 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 11. Vegetation classification for ROMO. GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Abies lasiocarpa - Abies lasiocarpa - Picea G5 CO, ID:S3, Subalpine Fir - CEGL000294 ROMO.707, ROMOAA.312 Picea engelmannii engelmannii / Acer MT, NM:S5, Engelmann Spruce / Forest Alliance glabrum Forest UT:S5, Rocky Mountain Maple WY:S3 Forest Abies lasiocarpa Abies lasiocarpa - Picea G5 CO:S5 Subalpine Fir - CEGL000296 ROMO.125, ROMO.292, ROMO.604, Seasonally Flooded engelmannii / Alnus Engelmann Spruce / ROMOAA.739 Forest Alliance incana Forest Speckled Alder Forest Abies lasiocarpa Abies lasiocarpa - Picea G5 AB, CO:S3, Subalpine Fir - CEGL000300 ROMO.050, ROMO.120, ROMO.288, ROMO.359, Seasonally Flooded engelmannii / ID:S3, Engelmann Spruce / ROMO.732, ROMOAA.1076, ROMOAA.380, Forest Alliance Calamagrostis MT:S5, Bluejoint Forest ROMOAA.442, ROMOAA.583, ROMOAA.913 canadensis Forest UT:S4?, WA?, WY:S2 Abies lasiocarpa Abies lasiocarpa - Picea G4G5 AZ:S2, CO, Subalpine Fir - CEGL000919 ROMO.397, ROMO.703, ROMOAA.357, Woodland Alliance engelmannii / Juniperus ID:S3?, Engelmann Spruce / ROMOAA.359, ROMOAA.509, ROMOAA.5209, communis Woodland MT:S3, Common Juniper ROMOAA.5216, ROMOAA.5221, NM?, NV?, Woodland ROMOAA.5243, ROMOAA.526, ROMOAA.5265, ROMOAA.5333, ROMOAA.5422, OR?, ROMOAA.5750, ROMOAA.5765, UT:S2?, ROMOAA.5767, ROMOAA.581, ROMOAA.582, WA:S3, ROMOAA.586, ROMOAA.648, ROMOAA.659, WY:S2? ROMOAA.795, ROMOAA.840, ROMOAA.904, ROMOAA.992 Abies lasiocarpa Abies lasiocarpa - Picea G5 CO:S5, Subalpine Fir - CEGL002663 ROMO.008, ROMO.170, ROMO.711, ROMO.816, Temporarily Flooded engelmannii / Mertensia NM:S5, Engelmann Spruce / ROMOAA.1154 Forest Alliance ciliata Forest WY? Mountain Bluebells Forest Abies lasiocarpa - Abies lasiocarpa - Picea G4 AZ:S2, Subalpine Fir - CEGL000321 ROMO.104, ROMO.118, ROMO.153, ROMO.431, Picea engelmannii engelmannii / Moss CO:S2?, Engelmann Spruce / ROMO.526, ROMOAA.1007, ROMOAA.15, Forest Alliance Forest NM:S4, Moss Forest ROMOAA.318, ROMOAA.5239, ROMOAA.5282, WY:S3 ROMOAA.5334, ROMOAA.5340, ROMOAA.5359, ROMOAA.5499, ROMOAA.5568, ROMOAA.584, ROMOAA.5901, ROMOAA.5910, ROMOAA.5916, ROMOAA.903, ROMOAA.919, ROMOAA.984

67 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Abies lasiocarpa - Abies lasiocarpa - Picea G5 CO:S4, Subalpine Fir - CEGL000373 ROMO.133, ROMO.702, ROMO.724, Picea engelmannii engelmannii / NM:S5 Engelmann Spruce / ROMOAA.459, ROMOAA.5854 Forest Alliance Polemonium Beautiful Jacob's-ladder pulcherrimum Forest Forest Abies lasiocarpa - Abies lasiocarpa - Picea GUQ CO:SU, UT? Subalpine Fir - CEGL000986 ROMOAA.109, ROMOAA.118, ROMOAA.119, Picea engelmannii - engelmannii / Salix Engelmann Spruce / ROMOAA.129, ROMOAA.136, ROMOAA.139, Pinus flexilis (brachycarpa, glauca) (Short-fruit Willow, ROMOAA.1128, ROMOAA.1135, Krummholz Shrubland Krummholz Shrubland Grayleaf Willow) ROMOAA.5094, ROMOAA.5106, ROMOAA.539, ROMOAA.5098, ROMOAA.5815, Alliance Krummholz Shrubland ROMOAA.5817, ROMOAA.5820, ROMOAA.5857 Abies lasiocarpa Abies lasiocarpa - Picea G5 CO:S4 Subalpine Fir - CEGL000327 ROMO.729, ROMO.731, ROMOAA.384, Seasonally Flooded engelmannii / Salix Engelmann Spruce / ROMOAA.813, ROMOAA.5177 Forest Alliance drummondiana Forest Drummond's Willow Forest Abies lasiocarpa - Abies lasiocarpa - Picea G5 CO:S4, Subalpine Fir - CEGL000340 ROMO.035, ROMO.063, ROMO.193, ROMO.235, Picea engelmannii engelmannii / ID:S3, Engelmann Spruce / ROMO.614, ROMOAA.100, ROMOAA.1160, Forest Alliance Vaccinium caespitosum MT:S5, Dwarf Blueberry Forest ROMOAA.135, ROMOAA.5845, ROMOAA.87, Forest UT:S4S5, ROMOAA.91 WA:S3? Abies lasiocarpa - Abies lasiocarpa - Picea G5 AZ:S2, Subalpine Fir - CEGL000343 ROMO.045, ROMO.324, ROMO.415, Picea engelmannii engelmannii / CO:S5, Engelmann Spruce / ROMOAA.1010, ROMOAA.1042, Forest Alliance Vaccinium myrtillus NM:S5, Whortleberry Forest ROMOAA.1051, ROMOAA.1075, Forest UT:S3S4 ROMOAA.1096, ROMOAA.1107, ROMOAA.1116, ROMOAA.138, ROMOAA.300, ROMOAA.372, ROMOAA.373, ROMOAA.434, ROMOAA.437, ROMOAA.456, ROMOAA.466, ROMOAA.5235, ROMOAA.5247, ROMOAA.5313, ROMOAA.5320, ROMOAA.5322, ROMOAA.5327, ROMOAA.5352, ROMOAA.537, ROMOAA.5462, ROMOAA.5470, ROMOAA.5473, ROMOAA.5487, ROMOAA.5504, ROMOAA.5505, ROMOAA.5510, ROMOAA.5751, ROMOAA.672, ROMOAA.680, ROMOAA.791, ROMOAA.792, ROMOAA.843, ROMOAA.845, ROMOAA.850, ROMOAA.852, ROMOAA.856, ROMOAA.857, ROMOAA.94, ROMOAA.988

68 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Abies lasiocarpa - Abies lasiocarpa - Picea G5 AZ?, CO:S5, Subalpine Fir - CEGL000344 ROMO.005, ROMO.018, ROMO.025, ROMO.027, Picea engelmannii engelmannii / ID:S5, Engelmann Spruce / ROMO.111, ROMO.124, ROMO.126, ROMO.129, Forest Alliance Vaccinium scoparium MT:S5, Grouseberry Forest ROMO.171, ROMO.225, ROMO.232, ROMO.263, Forest NM?, ROMO.309, ROMO.330, ROMO.407, ROMOAA.102, ROMOAA.1038, ROMOAA.1043, OR:S4, ROMOAA.1050, ROMOAA.1052, UT:S5, ROMOAA.1062, ROMOAA.1095, WA:S4, ROMOAA.1108, ROMOAA.1150, WY:S5 ROMOAA.1153, ROMOAA.1159, ROMOAA.126, ROMOAA.133, ROMOAA.285, ROMOAA.286, ROMOAA.289, ROMOAA.462, ROMOAA.463, ROMOAA.499, ROMOAA.5318, ROMOAA.5325, ROMOAA.5328, ROMOAA.5331, ROMOAA.541, ROMOAA.542, ROMOAA.543, ROMOAA.544, ROMOAA.5450, ROMOAA.5451, ROMOAA.5458, ROMOAA.5463, ROMOAA.5468, ROMOAA.5472, ROMOAA.5502, ROMOAA.574, ROMOAA.5822, ROMOAA.5846, ROMOAA.595, ROMOAA.844, ROMOAA.847, ROMOAA.85, ROMOAA.853, ROMOAA.997 Abies lasiocarpa - Abies lasiocarpa G4 AB, CO, Subalpine Fir CEGL000985 ROMOAA.82, ROMOAA.83, ROMOAA.89, Picea engelmannii - Krummholz Shrubland MT:S4, UT?, Krummholz Shrubland ROMOAA.92, ROMOAA.93, ROMOAA.94, Pinus flexilis WY ROMOAA.95, ROMOAA.102, ROMOAA.103, Krummholz Shrubland ROMOAA.106, ROMOAA.108, ROMOAA.110, ROMOAA.113, ROMOAA.121, ROMOAA.124, Alliance ROMOAA.126, ROMOAA.127, ROMOAA.128, ROMOAA.169, ROMOAA.217, ROMOAA.5818, ROMOAA.5822, ROMOAA.5825, ROMOAA.5829, ROMOAA.5838, ROMOAA.5845, ROMOAA.5847, ROMOAA.5861, ROMOAA.5989, Acer glabrum Acer glabrum Drainage G4? CO, MT:S4, Rocky Mountain Maple CEGL001062 ROMO.188, ROMO.224, ROMOAA.268, Temporarily Flooded Bottom Shrubland WY:S3 Drainage Bottom ROMOAA.8006 Shrubland Alliance Shrubland Alnus incana Alnus incana - Salix G3 CO:S3, Speckled Alder - CEGL002651 ROMO.084, ROMOAA.859 Seasonally Flooded (monticola, lucida, NM:S3?, (Mountain Willow, Shrubland Alliance ligulifolia) Shrubland WY? Whiplash Willow, Strapleaf Willow) Shrubland

69 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Alnus incana Alnus incana - Salix G3 CO:S3, Speckled Alder - CEGL002652 ROMO.418 Temporarily Flooded drummondiana NM?, WY Drummond's Willow Shrubland Alliance Shrubland Shrubland Alnus incana Alnus incana / G3 BC?, CA?, Speckled Alder / Field CEGL001146 ROMO.064 Seasonally Flooded Equisetum arvense CO:S3, Horsetail Shrubland Shrubland Alliance Shrubland ID:S3, OR:S2, UT:S2S3, WA:S3, WY Alnus incana Alnus incana / Mesic G3 CO:S3, ID, Speckled Alder / Mesic CEGL001148 ROMO.039, ROMO.119, ROMO.127, ROMO.179, Temporarily Flooded Graminoids Shrubland NV, Graminoids Shrubland ROMO.223, ROMO.302, ROMO.626, Shrubland Alliance UT:S2S3, ROMOAA.254, ROMOAA.287, ROMOAA.391, WY ROMOAA.5077 Aquilegia (caerulea, Aquilegia caerulea - GU CO:SU Colorado Blue CEGL001938 ROMO.081, ROMO.169, ROMO.429, flavescens) Sparsely Cirsium scopulorum Columbine - Alpine ROMOAA.5380 Vegetated Alliance Scree Sparse Vegetation Thistle Scree Sparse Vegetation Artemisia arctica Artemisia arctica ssp. GU CO:SU Boreal Sagebrush CEGL001848 ROMO.391, ROMO.423, ROMO.436, ROMO.709, Herbaceous Alliance arctica Herbaceous Herbaceous Vegetation ROMOAA.767 Vegetation Artemisia tridentata Artemisia tridentata ssp. GNR CO Mountain Big Sagebrush CEGL005827 ROMO.107, ROMO.284, ROMO.315, ROMO.341, ssp. vaseyana vaseyana - (Purshia - (Bitterbrush) / ROMO.347, ROMO.494, ROMO.632, ROMO.800, Shrubland Alliance tridentata) / Mountain Muhly - ROMO.803, ROMO.813, ROMO.825, ROMO.843, Muhlenbergia montana (Needle-and-Thread) ROMOAA.1068, ROMOAA.1105, ROMOAA.188, ROMOAA.408, ROMOAA.412, ROMOAA.5130, - (Hesperostipa comata Shrubland ROMOAA.5142, ROMOAA.5147, ssp. comata) Shrubland ROMOAA.5156, ROMOAA.5157, ROMOAA.5161, ROMOAA.5259, ROMOAA.5791, ROMOAA.716, ROMOAA.718, ROMOAA.726, ROMOAA.814, ROMOAA.816, ROMOAA.817, ROMOAA.818, ROMOAA.819, ROMOAA.821, ROMOAA.822, ROMOAA.824, ROMOAA.825, ROMOAA.944, ROMOAA.953

70 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Betula nana Seasonally Betula nana - Salix GNR CO Swamp Birch / Short- CEGL005828 ROMO.151, ROMO.427, ROMO.438, ROMO.453, Flooded Shrubland brachycarpa Shrubland fruit Willow Shrubland ROMO.476, ROMOAA.1152, ROMOAA.216, Alliance ROMOAA.263, ROMOAA.399, ROMOAA.484, ROMOAA.5037, ROMOAA.5089, ROMOAA.5090, ROMOAA.5100 Betula nana Seasonally Betula nana / Mesic G3G4 CO:S3, WY Swamp Birch - Mesic CEGL002653 ROMO.034, ROMO.155, ROMO.219, ROMO.242, Flooded Shrubland Forbs - Mesic Forbs - Mesic ROMO.310, ROMO.469, ROMO.726, ROMO.741, Alliance Graminoids Shrubland Graminoids Shrubland ROMOAA.1133, ROMOAA.232 Betula occidentalis Betula occidentalis / G3 CO:S2, NV, Water Birch / Mesic CEGL002654 ROMO.074, ROMO.123, ROMO.202, ROMO.486, Seasonally Flooded Mesic Graminoids UT Graminoids Shrubland ROMO.8003, ROMO.8008, ROMO.8010, Shrubland Alliance Shrubland ROMOAA.420, ROMOAA.5083, ROMOAA.5922, ROMOAA.714, ROMOAA.715, ROMOAA.8009, ROMOAA.8014, ROMOAA.830 Bouteloua gracilis Bouteloua gracilis G4Q AZ, CO, Blue Grama Herbaceous CEGL001760 ROMO.636, ROMOAA.179, ROMOAA.5134 Herbaceous Alliance Herbaceous Vegetation NM?, UT, Vegetation WY:S4 Bromus inermis Semi- Bromus inermis - GNA CO, MT, Smooth Brome - CEGL005264 ROMO.061, ROMOAA.159, ROMOAA.174, natural Herbaceous (Pascopyrum smithii) ND, SD, UT, (Western Wheatgrass) ROMOAA.195, Alliance Semi-natural WY Semi-natural Herbaceous Vegetation Herbaceous Vegetation

Calamagrostis Calamagrostis G4 AB, Bluejoint Western CEGL001599 ROMO.049, ROMO.115, ROMO.175, ROMO.267, canadensis Seasonally canadensis Western BC:S3S4, Herbaceous Vegetation ROMO.291, ROMO.482, ROMOAA.479, Flooded Herbaceous Herbaceous Vegetation CA, CO:S4, ROMOAA.592, ROMOAA.75 Alliance ID:S4, MT:S4, ND, OR:S3S4, SD, UT:S2S3, WA:S3S4, WY:S2 Caltha leptosepala Caltha leptosepala - GNRQ CO:SU White Marsh-marigold - CEGL001957 ROMO.231, ROMO.240 Saturated Herbaceous Rhodiola rhodantha Queen's Crown Alliance Herbaceous Vegetation Herbaceous Vegetation

71 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Caltha leptosepala Caltha leptosepala G4 CO:S4, White Marsh-marigold CEGL001954 ROMO.011, ROMO.037, ROMO.323, ROMO.332, Saturated Herbaceous Herbaceous Vegetation ID:S4, MT, Herbaceous Vegetation ROMO.433, ROMO.481 Alliance UT:S2S3, WY Cardamine cordifolia Cardamine cordifolia - G4 CO:S4 Large Mountain CEGL002662 ROMO.002, ROMO.233, ROMO.237, ROMO.241, Saturated Herbaceous Mertensia ciliata - Bittercress - Mountain ROMO.247, ROMO.252, ROMO.255, ROMO.366, Alliance Senecio triangularis Bluebells - Arrowleaf ROMO.399, ROMO.463, ROMO.725, Herbaceous Vegetation Ragwort Herbaceous ROMOAA.1136, ROMOAA.5885, ROMOAA.777 Vegetation Carex aquatilis Carex aquatilis - Carex G4 AB, CO:S4, Aquatic Sedge - Beaked CEGL001803 ROMO.038, ROMO.114, ROMOAA.230 Seasonally Flooded utriculata Herbaceous MT:S3, NB?, Sedge Herbaceous Herbaceous Alliance Vegetation WY? Vegetation Carex aquatilis Carex aquatilis G5 AB, AZ?, Aquatic Sedge CEGL001802 ROMO.003, ROMO.113, ROMO.147, ROMO.154, Seasonally Flooded Herbaceous Vegetation CA:S3, Herbaceous Vegetation ROMO.215, ROMO.226, ROMO.251, ROMO.254, Herbaceous Alliance CO:S4, ROMO.265, ROMO.333, ROMO.381, ROMO.419, ID:S4, ROMO.430, ROMO.457, ROMOAA.228, ROMOAA.233, ROMOAA.251, ROMOAA.5930, MT:S4, ROMOAA.607, ROMOAA.645, ROMOAA.69, NM:S4, NV, ROMOAA.80 OR:S4, UT:S3?, WA:S3, WY:S3 Carex elynoides Carex elynoides – G4 CO, MT:S4 Blackroot Sedge - Ross' CEGL001853 ROMO.456, ROMO.459, ROMOAA.475, Herbaceous Alliance Geum rossii Avens Herbaceous ROMOAA.5371, ROMOAA.5379, Herbaceous Vegetation Vegetation ROMOAA.5970, ROMOAA.6010 Carex (lachenalii, Carex illota Herbaceous GUQ CO:SU, Small-head Sedge CEGL001876 ROMO.141, ROMO.172, ROMO.270, ROMO.480, capillaris, illota) Vegetation OR:S3, WY Herbaceous Vegetation ROMOAA.1058, ROMOAA.5928, ROMOAA.5951 Seasonally Flooded Herbaceous Alliance

72 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Carex limosa Carex limosa G2 BC?, CA:S1, Mud Sedge Herbaceous CEGL001811 ROMO.824 Seasonally Flooded Herbaceous Vegetation CO, ID:S1, Vegetation Herbaceous Alliance MT:S2, NM, UT:S1S2, WA:S2?, WY:S2? Carex microptera Carex microptera G4 CO:S2?, ID?, Small-wing Sedge CEGL001792 ROMO.8018 Seasonally Flooded Herbaceous Vegetation OR?, Herbaceous Vegetation Herbaceous Alliance UT:S2S3, WY:S3 Carex pyrenaica Carex pyrenaica GU CO:SU Pyrenean Sedge CEGL001860 ROMO.007, ROMO.340, ROMO.390, ROMO.466, Herbaceous Alliance Herbaceous Vegetation Herbaceous Vegetation ROMO.474, ROMO.600 Carex rupestris Carex rupestris - Geum G4 CO:S4, UT, Curly Sedge - Ross' CEGL001861 ROMO.010, ROMO.030, ROMO.132, ROMO.137, Herbaceous Alliance rossii Herbaceous WY:S2 Avens Herbaceous ROMO.339, ROMO.355, ROMOAA111, Vegetation Vegetation ROMOAA.1047, ROMOAA.471, ROMOAA.5045, ROMOAA.550, ROMOAA.5600, ROMOAA.5602, ROMOAA.5620, ROMOAA.5982, ROMOAA.5997, ROMOAA.771, ROMOAA.979 Carex rupestris Carex rupestris - G3G4 CO:S3S4, Curly Sedge - Uinta CEGL001863 ROMO.165, ROMO.352, ROMO.385, ROMO.387, Herbaceous Alliance Trifolium dasyphyllum UT? Clover Herbaceous ROMO.439, ROMO.472, ROMOAA.224, Herbaceous Vegetation Vegetation ROMOAA.5028, ROMOAA.5615, ROMOAA.5976 Carex scopulorum Carex scopulorum - G4 CO:S4, Holm's Rocky Mountain CEGL001823 ROMO.052, ROMO.298, ROMO.468, ROMO.478, Seasonally Flooded Caltha leptosepala MT:S3, Sedge - White Marsh- ROMO.612, ROMOAA.1139, ROMOAA.1144, Herbaceous Alliance Herbaceous Vegetation WY:S2? marigold Herbaceous ROMOAA.225, ROMOAA.236, ROMOAA.247, Vegetation ROMOAA.772 Carex siccata Carex siccata - Geum GU CO:SU Dry-spike Sedge - Ross' CEGL001808 ROMO.458, ROMO.490, ROMOAA.155 Herbaceous Alliance rossii Herbaceous Avens Herbaceous Vegetation Vegetation

73 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Carex (rostrata, Carex utriculata G5 AB, AZ?, Beaked Sedge CEGL001562 ROMO.131, ROMO.152, ROMO.218, ROMO.229, utriculata) Seasonally Herbaceous Vegetation CA:S4, Herbaceous Vegetation ROMO.342, ROMO.367, ROMOAA.227, Flooded Herbaceous CO:S4, ROMOAA.270, ROMOAA.414, ROMOAA.63, Alliance ID:S4, ROMOAA.64, ROMOAA.67, ROMOAA.78 MT:S5, NM:S3, NV, OR:S4, UT:S3S4, WA:S3S4, WY:S3 Cercocarpus montanus Cercocarpus montanus / G2 CO:S2 Mountain-mahogany / CEGL001092 ROMOAA.558, ROMOAA.5652, ROMOAA.657 Shrubland Alliance Hesperostipa comata Needle-and-Thread Shrubland Shrubland Cercocarpus montanus Cercocarpus montanus / GU CO:S2 Mountain-mahogany / CEGL002914 ROMO.8007, ROMO.8017, ROMOAA.1161, Shrubland Alliance Muhlenbergia montana Mountain Muhly ROMOAA.614, ROMOAA.655, ROMOAA.656 Shrubland Shrubland Danthonia intermedia Danthonia intermedia G2G3 CO, Timber Oatgrass CEGL001794 ROMO.046, ROMO.070, ROMO.173, ROMO.176, Herbaceous Alliance Herbaceous Vegetation UT:S2S3, Herbaceous Vegetation ROMO.278, ROMO.283, ROMO.491, WA:S2? ROMOAA.158, ROMOAA.5835, ROMOAA.5926, ROMOAA.869 Danthonia parryi Danthonia parryi G3 CO:S3, Parry's Oatgrass CEGL001795 ROMO.489, ROMO.720 Herbaceous Alliance Herbaceous Vegetation WY:S2 Herbaceous Vegetation Dasiphora fruticosa Dasiphora fruticosa ssp. G4 CO:S3S4, Shrubby-cinquefoil / CEGL001107 ROMO.157, ROMO.194, ROMO.277, ROMO.740, Temporarily Flooded floribunda / ID:S3, Tufted Hairgrass ROMO.806, ROMOAA.411 Shrubland Alliance Deschampsia caespitosa MT:S4, Shrubland Shrubland OR:S2, UT:S3S4, WY:S3 Dasiphora fruticosa Dasiphora fruticosa ssp. G5? CO, NV:S5 Shrubby-cinquefoil CEGL001105 ROMO.117, ROMO.268, ROMO.814, Temporarily Flooded floribunda Shrubland Shrubland ROMOAA.409, ROMOAA.410, ROMOAA.413, Shrubland Alliance [Provisional] ROMOAA.621, ROMOAA.622, ROMOAA.713, ROMOAA.823 Dasiphora fruticosa Dasiphora fruticosa ssp. GNR CO, NM Shrubby-cinquefoil CEGL003499 ROMO.353, ROMO.452, ROMO.831, Temporarily Flooded floribunda Subalpine Subalpine Shrubland Shrubland Alliance Shrubland

74 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Deschampsia Deschampsia caespitosa G4 CO:S4, Tufted Hairgrass - White CEGL001882 ROMO.140, ROMO.264, ROMO.272, ROMO.475, caespitosa Saturated - Caltha leptosepala ID:S2, Marsh-marigold ROMOAA.1056, ROMOAA.1057, Herbaceous Alliance Herbaceous Vegetation MT:S3, WY Herbaceous Vegetation ROMOAA.1060, ROMOAA.476, ROMOAA.5950 Deschampsia Deschampsia caespitosa G5 CO:S5, Tufted Hairgrass - Ross' CEGL001884 ROMO.023, ROMO.083, ROMO.144, ROMO.248, caespitosa Temporarily - Geum rossii WY:S3? Avens Herbaceous ROMO.250, ROMO.331, ROMO.358, ROMO.444, Flooded Herbaceous Herbaceous Vegetation Vegetation ROMOAA.1131, ROMOAA.221, ROMOAA.237, Alliance ROMOAA.468, ROMOAA.5012, ROMOAA.547, ROMOAA.5603, ROMOAA.5609, ROMOAA.5984, ROMOAA.6000, ROMOAA.651, ROMOAA.660 Deschampsia Deschampsia caespitosa G4 AB, AZ:S2?, Tufted Hairgrass CEGL001599 ROMO.006, ROMO.036, ROMO.055, ROMO.139, caespitosa Seasonally Herbaceous Vegetation CA?, CO:S4, Herbaceous Vegetation ROMO.180, ROMO.190, ROMO.245, ROMO.253, Flooded Herbaceous ID:S3, ROMO.349, ROMO.377, ROMO.378, ROMO.447, Alliance MT:S4, NM, ROMO.451, ROMO.713, ROMOAA.1049, ROMOAA.1059, ROMOAA.1122, NV?, OR:S2, ROMOAA.1134, ROMOAA.1146, ROMOAA.144, UT:S3S4, ROMOAA.478, ROMOAA.5026, ROMOAA.5035, WA, WY ROMOAA.591, ROMOAA.594, ROMOAA.5944, ROMOAA.5946, ROMOAA.5964, ROMOAA.5968, ROMOAA.712, ROMOAA.742, ROMOAA.766, ROMOAA.781, ROMOAA.809, ROMOAA.86, ROMOAA.960 Dryas octopetala Dryas octopetala - G4 AB, CO:S4?, Eight-petal Mountain- CEGL001892 ROMO.150, ROMO.238, ROMO.246, ROMO.336, Dwarf-shrub Carex rupestris Dwarf- ID?, MT:S3 avens - Curly Sedge ROMO.356, ROMO.428, ROMO.441, ROMO.449, Herbaceous Alliance shrub Herbaceous Dwarf-shrub Herbaceous ROMO.506, ROMOAA.1083, ROMOAA.5013, Vegetation Vegetation ROMOAA.5605, ROMOAA.5963, ROMOAA.5978 Eleocharis Eleocharis quinqueflora G4 CA, Few-flower Spikerush CEGL001836 ROMO.004, ROMO.020, ROMO.021, ROMO.024, (quinqueflora, Herbaceous Vegetation CO:S3S4, Herbaceous Vegetation ROMO.048, ROMO.056, ROMO.072, ROMO.156, rostellata) Saturated ID:S1, ROMO.163, ROMO.227, ROMO.234, ROMO.296, Herbaceous Alliance MT:S3, ROMO.343, ROMO.375, ROMO.414, ROMO.421, ROMO.479, ROMO.493, ROMOAA.226, NV?, OR:S4, ROMOAA.231, ROMOAA.5949 UT:S2?, WA:S2?, WY:S2 na Elymus trachycaulus na CO Slender Wild Rye Local Type ROMO.809 Alliance [Placeholder] Alliance [Placeholder]

75 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Festuca brachyphylla Festuca brachyphylla - GUQ CO, NM:SU Shortleaf Fescue - Ross' CEGL001895 ROMO.426, ROMOAA.775 Herbaceous Alliance Geum rossii var. Avens Herbaceous turbinatum Herbaceous Vegetation Vegetation Festuca thurberi Festuca thurberi G3 CO, NM:S3 Thurber's Fescue CEGL001631 ROMO.635, ROMO.832 Herbaceous Alliance Subalpine Grassland Subalpine Grassland Herbaceous Vegetation Herbaceous Vegetation Geum rossii Geum rossii - Minuartia G3? CO, MT:S3 Ross' Avens - Alpine CEGL001965 ROMO.149, ROMO.443, ROMO.504, Herbaceous Alliance obtusiloba Herbaceous Stitchwort Herbaceous ROMOAA.206, ROMOAA.215, ROMOAA.5007, Vegetation Vegetation ROMOAA.5623, ROMOAA.5990 Geum rossii Geum rossii - G4G5 CO:S4S5, Ross' Avens - American CEGL001967 ROMO.715, ROMOAA.210, ROMOAA.5029, Herbaceous Alliance Polygonum bistortoides WY Bistort Herbaceous ROMOAA.708 Herbaceous Vegetation Vegetation Geum rossii Geum rossii - Sibbaldia GU CO:SU Ross' Avens - Creeping CEGL001969 ROMO.012, ROMO.192, ROMO.392, ROMO.467, Herbaceous Alliance procumbens Glow-wort Herbaceous ROMO.701, ROMOAA.166, ROMOAA.208, Herbaceous Vegetation Vegetation ROMOAA.241, ROMOAA.773 Geum rossii Geum rossii - Trifolium G3 CO, WY:S3 Ross' Avens - Clover CEGL001970 ROMO.009, ROMO.082, ROMO.085, ROMO.189, Herbaceous Alliance spp. Herbaceous species Herbaceous ROMO.345, ROMO.379, ROMO.384, ROMO.465, Vegetation Vegetation ROMO.477, ROMO.492, ROMOAA.220, ROMOAA.5047, ROMOAA.5387, ROMOAA.5613, ROMOAA.5626, ROMOAA.707, ROMOAA.776 Glyceria (grandis, Glyceria grandis G2? CO, ID:S1?, American Mannagrass CEGL003429 ROMO.637, ROMOAA.79 striata) Seasonally Herbaceous Vegetation MT, Herbaceous Vegetation Flooded Herbaceous OR:S1S2, Alliance WA:S1S2

76 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Juncus balticus Juncus balticus G5 BC:S3, CA?, Baltic Rush Herbaceous CEGL001838 ROMO.100, ROMO.121, ROMO.222, ROMO.314, Seasonally Flooded Herbaceous Vegetation CO:S5, Vegetation ROMO.338, ROMO.807, ROMOAA.5067, Herbaceous Alliance ID:S5, ROMOAA.578, ROMOAA.5942, ROMOAA.710, MT:S5, NE, ROMOAA.711 NM:S4, NV, OR:S5, SD, UT:S3S4, WA:S3S4, WY:S3 Juncus drummondii Juncus drummondii - G4 CO:SU, Drummond's Rush - CEGL001905 ROMO.043, ROMO.257, ROMO.484, ROMO.601, Herbaceous Alliance Carex spp. Herbaceous ID:S4 Sedge species ROMO.823, ROMOAA.5617, ROMOAA.769, Vegetation Herbaceous Vegetation ROMOAA.770 Juncus parryi Juncus parryi / G3G4 AB, CO, Parry's Rush / Creeping CEGL005871 ROMO.068, ROMO.164, ROMO.177, ROMO.191, Herbaceous Alliance Sibbaldia procumbens MT:S3?, WY Glow-wort Herbaceous ROMO.294, ROMO.299 Herbaceous Vegetation Vegetation Juniperus scopulorum Juniperus scopulorum / G2 CO:S2, WY? Rocky Mountain Juniper CEGL000749 ROMO.402, ROMO.623, ROMO.628, ROMO.631, Woodland Alliance Purshia tridentata / Bitterbrush Woodland ROMO.828, ROMOAA.326, ROMOAA.328, Woodland ROMOAA.520, ROMOAA.5201, ROMOAA.5638, ROMOAA.5640, ROMOAA.5645, ROMOAA.5647, ROMOAA.5651, ROMOAA.930, ROMOAA.936, ROMOAA.973 Kobresia myosuroides Kobresia myosuroides - G3 CO:S3 Pacific Bog Sedge - CEGL001907 ROMO.344, ROMO.442, ROMOAA.5004 Herbaceous Alliance Carex rupestris var. Drummond's Curly drummondiana Sedge Herbaceous Herbaceous Vegetation Vegetation Kobresia myosuroides Kobresia myosuroides - G5 CO:S5, Pacific Bog Sedge - CEGL001908 ROMO.166, ROMO.450, ROMO.460, ROMO.505, Herbaceous Alliance Geum rossii NM:S4, Ross' Avens Herbaceous ROMOAA.218, ROMOAA.222, ROMOAA.223, Herbaceous Vegetation UT:S4S5 Vegetation ROMOAA.242, ROMOAA.546, ROMOAA.5612 Muhlenbergia montana Muhlenbergia montana G1G2 CO:S2? Mountain Muhly - CEGL001647 ROMO.487, ROMO.804, ROMO.808, ROMO.834, Herbaceous Alliance - Hesperostipa comata Needle-and-Thread ROMO.839, ROMOAA.176, ROMOAA.5783, Herbaceous Vegetation Herbaceous Vegetation ROMOAA.5785, ROMOAA.719

77 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Muhlenbergia montana Muhlenbergia montana G3G4 AZ, CO:S2?, Mountain Muhly CEGL001646 ROMO.041, ROMO.835, ROMOAA.178, Herbaceous Alliance Herbaceous Vegetation UT:S3S4 Herbaceous Vegetation ROMOAA.5132, ROMOAA.5784, ROMOAA.5800, ROMOAA.674, ROMOAA.731 Nymphaea odorata - Nuphar lutea ssp. G5 BC:S5, CA, Yellow Pond-lily CEGL002001 ROMO.102, ROMO.322, ROMO.416, ROMO.728, Nuphar spp. polysepala Herbaceous CO:S3, Herbaceous Vegetation ROMOAA.76 Permanently Flooded Vegetation ID:S4, MT, Temperate Herbaceous OR:S5, Alliance WA:S4S5 Paronychia pulvinata Paronychia pulvinata - G5 AZ, CO:S5, Rocky Mountain CEGL001976 ROMO.357, ROMO.613, ROMO.716, ROMO.721, Dwarf-shrubland Silene acaulis Dwarf- NM?, Nailwort - Cushion Pink ROMOAA.145 Alliance shrubland UT:S4S5, Dwarf-shrubland WY Picea engelmannii Picea engelmannii / G4 CO, ID:S4, Engelmann Spruce / CEGL002678 ROMO.346, ROMO.605, Saturated Forest Calamagrostis MT:S4, Bluejoint Forest ROMO.607ROMOAA.1074, ROMOAA.5269, Alliance canadensis Forest WY:S3 ROMOAA.950 Picea engelmannii Picea engelmannii / G4 AB, CO:S2, Engelmann Spruce / CEGL005927 ROMO.216, ROMO.325, ROMOAA.448 Seasonally Flooded Equisetum arvense ID:S2, Field Horsetail Forest Forest Alliance Forest MT:S4, OR:S3, UT:S3?, WA:S3, WY:S2 Picea engelmannii Picea engelmannii / G4 AZ:S1, Engelmann Spruce / CEGL000371 ROMO.182, ROMO.266, ROMO.326, Forest Alliance Moss Forest CO:SU, Moss Forest ROMOAA.849 NM:S4 Picea engelmannii Picea engelmannii / G2? CO:S2 Engelmann Spruce / CEGL000377 ROMO.199, ROMOAA.82, ROMOAA.92, Forest Alliance Trifolium dasyphyllum Uinta Clover Forest ROMOAA.5844 Forest Picea engelmannii Picea engelmannii / G4Q CO, NM:S4 Engelmann Spruce / CEGL000379 ROMO.207, ROMO.370, ROMOAA.255, Forest Alliance Vaccinium myrtillus Whortleberry Forest ROMOAA.288, ROMOAA.316, ROMOAA.5326, Forest ROMOAA.5482, ROMOAA.5484, ROMOAA.5491

78 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Picea engelmannii Picea engelmannii / G3G5 CO, MT?, Engelmann Spruce / CEGL000381 ROMO.815, ROMOAA.1151, ROMOAA.302, Forest Alliance Vaccinium scoparium NM, OR:SU, Grouseberry Forest ROMOAA.5321, ROMOAA.5323, Forest UT:S4S5, ROMOAA.5344, ROMOAA.5466, WY:S3 ROMOAA.5475, ROMOAA.5506, ROMOAA.5578, ROMOAA.573, ROMOAA.81, ROMOAA.810, ROMOAA.98 Picea pungens Picea pungens / Alnus G3 CO:S3, Blue Spruce / Speckled CEGL000894 ROMO.184, ROMO.827, ROMOAA.5913, Temporarily Flooded incana Woodland NM?, WY Alder Woodland ROMOAA.5921, ROMOAA.735, ROMOAA.925 Woodland Alliance Pinus contorta Forest Pinus contorta / G5 CO, MT?, Lodgepole Pine / CEGL000134 ROMO.174, ROMO.311, ROMO.313, ROMO.818, Alliance Arctostaphylos uva-ursi OR:S4, Kinikinnick Forest ROMOAA.1102, ROMOAA.306, ROMOAA.343, Forest UT:S4S5, ROMOAA.35, ROMOAA.421, ROMOAA.511, WA:S4?, ROMOAA.5248, ROMOAA.525, ROMOAA.5262, ROMOAA.5419, ROMOAA.5516, WY:S3 ROMOAA.5554, ROMOAA.658, ROMOAA.875 Pinus contorta Forest Pinus contorta / Carex G4? CO:S4, Lodgepole Pine / Geyer's CEGL000141 ROMO.134, ROMO.327, ROMOAA.388, Alliance geyeri Forest ID:S4, MT?, Sedge Forest ROMOAA.439 OR?, WY:S3S4 Pinus contorta Forest Pinus contorta / Carex G5 CO, ID, Lodgepole Pine / Ross' CEGL000144 ROMO.8001, ROMO.812 Alliance rossii Forest MT?, Sedge Forest UT:S4?, WY:S5 Pinus contorta Forest Pinus contorta / Jamesia GNR CO Lodgepole Pine / CEGL005933 ROMO.060, ROMO.210, ROMO.512, ROMO.517, Alliance americana Forest Waxflower Forest ROMO.743, ROMOAA.291, ROMOAA.5256, ROMOAA.5399, ROMOAA.5541, ROMOAA.5567, ROMOAA.650, ROMOAA.879, ROMOAA.883

79 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pinus contorta Pinus contorta / G5 CO:S3, Lodgepole Pine / CEGL000764 ROMO.106, ROMO.204, ROMO.306, Woodland Alliance Juniperus communis ID:S3, Common Juniper ROMOAA.1004, ROMOAA.1026, Woodland MT:S3, Woodland ROMOAA.1034, ROMOAA.1070, NV?, ROMOAA.1086, ROMOAA.1091, ROMOAA.192, ROMOAA.308, ROMOAA.310, ROMOAA.315, UT:S4?, ROMOAA.332, ROMOAA.338, ROMOAA.387, WY:S3 ROMOAA.389, ROMOAA.424, ROMOAA.494, ROMOAA.495, ROMOAA.497, ROMOAA.500, ROMOAA.517, ROMOAA.5213, ROMOAA.5233, ROMOAA.5275, ROMOAA.5306, ROMOAA.5418, ROMOAA.5447, ROMOAA.5517, ROMOAA.5544, ROMOAA.5563, ROMOAA.5564, ROMOAA.5574, ROMOAA.5768, ROMOAA.59, ROMOAA.695, ROMOAA.782, ROMOAA.788, ROMOAA.826, ROMOAA.899, ROMOAA.902, ROMOAA.929, ROMOAA.993 Pinus contorta Pinus contorta / Rock GNR CO Lodgepole Pine / Rock CEGL005934 ROMO.128, ROMO.183, ROMO.289, Woodland Alliance Woodland Woodland ROMOAA.1001, ROMOAA.1002, ROMOAA.1013, ROMOAA.1033, ROMOAA.1039, ROMOAA.1066, ROMOAA.13, ROMOAA.313, ROMOAA.319, ROMOAA.356, ROMOAA.36, ROMOAA.5224, ROMOAA.524, ROMOAA.5403, ROMOAA.5410, ROMOAA.563, ROMOAA.585, ROMOAA.611, ROMOAA.793, ROMOAA.797, ROMOAA.8, ROMOAA.876, ROMOAA.882, ROMOAA.9, ROMOAA.946 Pinus contorta Forest Pinus contorta / G3G4 CO:S3S4, Lodgepole Pine / Russet CEGL000163 ROMO.105, ROMO.130, ROMO.348, Alliance Shepherdia canadensis ID, MT?, Buffaloberry Forest ROMOAA.311, ROMOAA.789 Forest WA:S3?, WY:S3S4 Pinus contorta Forest Pinus contorta / G5 AB, CO, Lodgepole Pine / Dwarf CEGL000168 ROMO.158, ROMO.196, ROMO.276, Alliance Vaccinium caespitosum ID:S4?, Blueberry Forest ROMOAA.608 Forest MT:S5, OR:S3, UT

80 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pinus contorta Forest Pinus contorta / GNR CO Lodgepole Pine / CEGL005935 ROMO.109, ROMO.205, ROMO.411, ROMO.417, Alliance Vaccinium myrtillus Whortleberry Forest ROMOAA.1064, ROMOAA.1078, Forest ROMOAA.1090, ROMOAA.1094, ROMOAA.1157, ROMOAA.299, ROMOAA.5485, ROMOAA.5534, ROMOAA.5537, ROMOAA.5588, ROMOAA.5594, ROMOAA.796, ROMOAA.798, ROMOAA.966 Pinus contorta Forest Pinus contorta / G5 CA?, CO:S4, Lodgepole Pine / CEGL000172 ROMO.304, ROMO.305, ROMO.308, Alliance Vaccinium scoparium ID:S5, Grouseberry Forest ROMOAA.1063, ROMOAA.298, ROMOAA.303, Forest MT:S5, ROMOAA.309, ROMOAA.457, ROMOAA.5461, OR:S3, ROMOAA.5486, ROMOAA.5524, ROMOAA.5535, ROMOAA.600, ROMOAA.703, UT:S4S5, ROMOAA.848, ROMOAA.918 WA:S4, WY:S5 Pinus flexilis Pinus flexilis / G4 AB, CO:S2?, Limber Pine / CEGL000802 ROMO.318, ROMO.371, ROMOAA.5253 Woodland Alliance Arctostaphylos uva-ursi MT, NM:S4 Kinikinnick Woodland Woodland Pinus flexilis Pinus flexilis / G5 CA?, CO:S3, Limber Pine / Common CEGL000807 ROMO.103, ROMO.236, ROMO.303, ROMO.317, Woodland Alliance Juniperus communis ID:S3, Juniper Woodland ROMO.369, ROMO.507, ROMO.518, ROMO.527, Woodland MT:S4, ROMO.617, ROMOAA.11, ROMOAA.342, NV?, OR:S1, ROMOAA.358, ROMOAA.465, ROMOAA.508, ROMOAA.540, ROMOAA.5430, ROMOAA.5448, UT, WY:S2? ROMOAA.5460, ROMOAA.5762, ROMOAA.5764, ROMOAA.5770, ROMOAA.5776, ROMOAA.5777, ROMOAA.662, ROMOAA.678, ROMOAA.753, ROMOAA.807 na Pinus flexilis na CO Limber Pine Krummholz Local Type ROMOAA.355, ROMOAA.394, ROMOAA.5851, Krummholz Shrubland Shrubland ROMOAA.5852, ROMOAA.5859 Pinus ponderosa Pinus ponderosa / G4 CO:S3, Ponderosa Pine / CEGL000844 ROMO.372, ROMO.412, ROMO.616, ROMO.706, Woodland Alliance Arctostaphylos uva-ursi MT:S3, Kinikinnick Woodland ROMOAA.1071, ROMOAA.204, ROMOAA.345, Woodland NM:S4, ROMOAA.368, ROMOAA.453, ROMOAA.5182, SD:S4, UT?, ROMOAA.52, ROMOAA.53, ROMOAA.5406, ROMOAA.5437, ROMOAA.5635, WY:S3 ROMOAA.5675, ROMOAA.5677, ROMOAA.5700, ROMOAA.5711, ROMOAA.5714, ROMOAA.641, ROMOAA.701, ROMOAA.803, ROMOAA.890, ROMOAA.975

81 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pinus ponderosa Pinus ponderosa / GNR CO, UT Ponderosa Pine / CEGL002794 ROMO.186, ROMO.488, ROMOAA.269, Woodland Alliance Artemisia tridentata ssp. Mountain Big Sagebrush ROMOAA.350, ROMOAA.353, ROMOAA.948 vaseyana Woodland Woodland Pinus ponderosa Pinus ponderosa / Carex G3G4 CO:S2S3, Ponderosa Pine / Geyer's CEGL000182 ROMO.290, ROMO.837, ROMOAA.5205, Woodland Alliance geyeri Woodland MT?, OR:S2, Sedge Woodland ROMOAA.5660 UT:S4?, WY:S3 Pinus ponderosa Pinus ponderosa / Carex G3G4 CO:S2, Ponderosa Pine / Sun CEGL000849 ROMOAA.390, ROMOAA.43, ROMOAA.5688, Woodland Alliance inops ssp. heliophila MT:S3S4, Sedge Woodland ROMOAA.928 Woodland SD, WY:S2S3 Pinus ponderosa Forest Pinus ponderosa / Carex G4G5 CO:S3S4, Ponderosa Pine / Ross' CEGL000183 ROMO.373, ROMO.801, ROMO.805, Alliance rossii Forest SD, WY:S3 Sedge Forest ROMOAA.1036, ROMOAA.5739, ROMOAA.737 Pinus ponderosa Pinus ponderosa / G4 CO:S4, Ponderosa Pine / CEGL000851 ROMO.8016 Woodland Alliance Cercocarpus montanus NM?, WY? Mountain-mahogany Woodland Woodland Pinus ponderosa Pinus ponderosa / G4? CO, MT:S3, Ponderosa Pine / CEGL000859 ROMOAA.1085, ROMOAA.1120, ROMOAA.17, Woodland Alliance Juniperus communis SD:S4, Common Juniper ROMOAA.175, ROMOAA.294, ROMOAA.337, Woodland WY:S3? Woodland ROMOAA.440, ROMOAA.5241, ROMOAA.5258, ROMOAA.5303, ROMOAA.5401, ROMOAA.5411, ROMOAA.5412, ROMOAA.5428, ROMOAA.5440, ROMOAA.5561, ROMOAA.5712, ROMOAA.6, ROMOAA.666, ROMOAA.736, ROMOAA.746, ROMOAA.786, ROMOAA.832, ROMOAA.892 Pinus ponderosa Pinus ponderosa / G3 CO:S3, Ponderosa Pine / Spike CEGL000186 ROMO.845, ROMOAA.5436, ROMOAA.5666, Woodland Alliance Leucopoa kingii WY:SU Fescue Woodland ROMOAA.5678, ROMOAA.633, ROMOAA.638 Woodland

82 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pinus ponderosa Pinus ponderosa / G4G5 AZ:S4, Ponderosa Pine / CEGL000862 ROMO.300, ROMO.510, ROMO.521, Woodland Alliance Muhlenbergia montana CO:S2S3, Mountain Muhly ROMOAA.1077, ROMOAA.1118, Woodland NM:S4, TX, Woodland ROMOAA.1121, ROMOAA.344, ROMOAA.352, UT:S4S5 ROMOAA.382, ROMOAA.39, ROMOAA.504, ROMOAA.507, ROMOAA.5187, ROMOAA.5190, ROMOAA.5416, ROMOAA.559, ROMOAA.5649, ROMOAA.5685, ROMOAA.5718, ROMOAA.5728, ROMOAA.5731, ROMOAA.5787, ROMOAA.5807, ROMOAA.675, ROMOAA.748, ROMOAA.871, ROMOAA.889 Pinus ponderosa Forest Pinus ponderosa / G3 CO, SD:S4, Ponderosa Pine / CEGL000190 ROMO.413, ROMO.523, ROMOAA.5661 Alliance Physocarpus WY:S2? Mountain Ninebark monogynus Forest Forest Pinus ponderosa Pinus ponderosa / G3G5 CA:S2, Ponderosa Pine / CEGL000867 ROMO.042, ROMO.075, ROMO.209, ROMO.401, Woodland Alliance Purshia tridentata CO:S3?, Bitterbrush Woodland ROMOAA.1109, ROMOAA.2, ROMOAA.320, Woodland ID:S3, ROMOAA.321, ROMOAA.354, ROMOAA.363, MT:S3, ROMOAA.419, ROMOAA.5118, ROMOAA.5120, ROMOAA.5124, ROMOAA.5128, UT:S5 ROMOAA.5139, ROMOAA.5169, ROMOAA.5170, ROMOAA.55, ROMOAA.554, ROMOAA.560, ROMOAA.5650, ROMOAA.5655, ROMOAA.5658, ROMOAA.5662, ROMOAA.5667, ROMOAA.5679, ROMOAA.5691, ROMOAA.5694, ROMOAA.5699, ROMOAA.5702, ROMOAA.5727, ROMOAA.5740, ROMOAA.5747, ROMOAA.637, ROMOAA.65, ROMOAA.720, ROMOAA.727, ROMOAA.743, ROMOAA.947, ROMOAA.956 Pinus ponderosa Forest Pinus ponderosa / Ribes GNR CO:SU Ponderosa Pine / White CEGL000199 ROMO.522, ROMO.610, ROMO.705, Alliance cereum Forest Squaw Currant Forest ROMOAA.1165, ROMOAA.22, ROMOAA.335, ROMOAA.42, ROMOAA.518, ROMOAA.5341, ROMOAA.5439, ROMOAA.725, ROMOAA.923

83 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pinus ponderosa Pinus ponderosa / G5? AZ, CO, Ponderosa Pine / CEGL000877 ROMO.201, ROMO.220, ROMO.319, ROMO.403, Woodland Alliance Rockland Woodland NM:S5, UT Rockland Woodland ROMOAA.1087, ROMOAA.330, ROMOAA.347, ROMOAA.38, ROMOAA.46, ROMOAA.493, ROMOAA.512, ROMOAA.5203, ROMOAA.5569, ROMOAA.5690, ROMOAA.5701, ROMOAA.605, ROMOAA.634, ROMOAA.636, ROMOAA.868, ROMOAA.878, ROMOAA.893 na Poa cusickii / Sibbaldia na CO Cusick's Bluegrass - Local Type ROMO.293 procumbens Creeping Glow-wort Herbaceous Vegetation Herbaceous Vegetation Populus angustifolia Populus angustifolia / G3 CO:S2, Narrowleaf Cottonwood CEGL000648 ROMO.8011, ROMO.8015 Temporarily Flooded Betula occidentalis ID:S1, / Water Birch Woodland Woodland Alliance Woodland NV:S2, UT:S3?, WY:S2 Populus angustifolia Populus angustifolia / GNR CO Narrowleaf Cottonwood CEGL005847 ROMO.841, ROMOAA.5629 Temporarily Flooded Bromus inermis Semi- / Smooth Brome Semi- Woodland Alliance natural Woodland natural Woodland na Populus balsamifera / na CO Balsam Poplar / Local Type ROMO.633 Pteridium aquilinum Northern Bracken Woodland Woodland Abies lasiocarpa - Populus tremuloides - G3G4 CO:S4, Quaking Aspen - CEGL000527 ROMO.112, ROMO.838, ROMOAA.3, Populus tremuloides Abies lasiocarpa / UT:S3S4 Subalpine Fir / Common ROMOAA.4, ROMOAA.489, ROMOAA.5250, Forest Alliance Juniperus communis Juniper Forest ROMOAA.587, ROMOAA.754, ROMOAA.915 Forest Picea pungens - Populus tremuloides - G3G4 CO, Quaking Aspen - Blue CEGL000535 ROMO.833, ROMOAA.1112, ROMOAA.274 Populus tremuloides Picea pungens Forest UT:S3S4 Spruce Forest Forest Alliance Pinus contorta - Populus tremuloides - G4G5 UT:S4S5, Quaking Aspen - CEGL000537 ROMOAA.1, ROMOAA.258, ROMOAA.431, Populus tremuloides Pinus contorta / WY Lodgepole Pine / ROMOAA.438, ROMOAA.447, ROMOAA.5208, Forest Alliance Juniperus communis Common Juniper Forest ROMOAA.5268, ROMOAA.5287, ROMOAA.794, Forest ROMOAA.836, ROMOAA.1065, ROMOAA.1069, ROMOAA.1099,ROMOAA.1016

84 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pinus ponderosa - Populus tremuloides - G3G4 CO:S3S4, Quaking Aspen - CEGL000541 ROMOAA.1012, ROMOAA.1035, ROMOAA.426, Populus tremuloides Pinus ponderosa Rocky UT:S3S4, Ponderosa Pine Rocky ROMOAA.436, ROMOAA.5184, ROMOAA.5189, Forest Alliance Mountain Forest WY? Mountain Forest ROMOAA.5345, ROMOAA.5425, ROMOAA.562, ROMOAA.722, ROMOAA.911, ROMOAA.952, ROMOAA.976 Populus tremuloides - Populus tremuloides - G3G4 UT:S3S4 Quaking Aspen - CEGL000545 ROMO.101, ROMOAA.336, ROMOAA.428, Pseudotsuga menziesii Pseudotsuga menziesii / Douglas-fir / Common ROMOAA.5171, ROMOAA.5231, Forest Alliance Juniperus communis Juniper Forest ROMOAA.5249, ROMOAA.5251, Forest ROMOAA.5875, ROMOAA.690, ROMOAA.835, ROMOAA.877, Populus tremuloides Populus tremuloides / G1G2 CO:S1S2 Quaking Aspen / Rocky CEGL000563 ROMO.122, ROMO.485, ROMO.708, Forest Alliance Acer glabrum Forest Mountain Maple Forest ROMOAA.1164, ROMOAA.24, ROMOAA.5586, ROMOAA.961 Populus tremuloides Populus tremuloides / G3 CO:S3 Quaking Aspen / CEGL001150 ROMO.033, ROMO.076, ROMO.217, Temporarily Flooded Alnus incana Forest Speckled Alder Forest ROMOAA.279, ROMOAA.1089, ROMOAA.282, Forest Alliance ROMOAA.283, ROMOAA.444, ROMOAA.5172, ROMOAA.618, ROMOAA.686, ROMOAA.942 Populus tremuloides Populus tremuloides / G3 AB, CO:S3, Quaking Aspen / CEGL000574 ROMO.069, ROMO.520, ROMO.630, ROMO.817, Seasonally Flooded Calamagrostis ID:S2, Bluejoint Forest ROMOAA.277, ROMOAA.381, ROMOAA.5308 Forest Alliance canadensis Forest MT:S2, OR:S1, WA:S1, WY? Populus tremuloides Populus tremuloides / G4 CO:S4, MT?, Quaking Aspen / CEGL000587 ROMO.135, ROMO.197, ROMO.704, ROMO.733, Forest Alliance Juniperus communis UT, WY:S3 Common Juniper Forest ROMO.810, ROMOAA.1061, ROMOAA.1092, Forest ROMOAA.256, ROMOAA.264, ROMOAA.365, ROMOAA.366, ROMOAA.425, ROMOAA.5299, ROMOAA.5725, ROMOAA.5872, ROMOAA.780, ROMOAA.842, ROMOAA.867, ROMOAA.901, ROMOAA.963 Populus tremuloides Populus tremuloides / GNR CO Quaking Aspen/ CEGL005829 ROMO.067, ROMO.502, ROMO.503, ROMO.609, Temporarily Flooded Phleum pratense Semi- Timothy Semi-natural ROMO.627, ROMO.744, ROMO.842, Forest Alliance natural Forest Forest ROMOAA.1000, ROMOAA.423, ROMOAA.5175, ROMOAA.5294, ROMOAA.5305, ROMOAA.863

85 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Populus tremuloides Populus tremuloides / GNR CO Quaking Aspen / CEGL005932 ROMO.509, ROMO.629, ROMOAA.281, Forest Alliance Physocarpus Mountain Ninebark ROMOAA.490, monogynus Forest Forest Populus tremuloides Populus tremuloides / GNR CA, CO, NV Quaking Aspen / CEGL003148 ROMO.519, ROMO.528, ROMOAA.193, Forest Alliance Poa pratensis Forest Kentucky Bluegrass ROMOAA.272, ROMOAA.276, ROMOAA.5173, Forest ROMOAA.5181, ROMOAA.747 Populus tremuloides Populus tremuloides / G3G4 CO, MT?, Quaking Aspen / Choke CEGL000596 ROMO.621, ROMOAA.698 Forest Alliance Prunus virginiana OR:SU, Cherry Forest Forest SD:S3, UT, WY:S2S3 Populus tremuloides Populus tremuloides / G4 CO:S3S4, Quaking Aspen / CEGL000597 ROMO.406 Forest Alliance Pteridium aquilinum SD, Northern Bracken Forest Forest UT:S2S3, WY Populus tremuloides Populus tremuloides / G3G4 CO:S3?, Quaking Aspen / Russet CEGL000606 ROMO.730 Forest Alliance Shepherdia canadensis ID:S2S3, Buffaloberry Forest Forest WY:S2S3 Populus tremuloides Populus tremuloides / G5 CA?, CO:S5, Quaking Aspen / CEGL000619 ROMO.110, ROMO.198, ROMO.511 Forest Alliance Thalictrum fendleri ID:S3, Fendler's Meadowrue Forest UT:S4S5, Forest WY:S3 Populus tremuloides Populus tremuloides / G3 CO:S3 Quaking Aspen / CEGL000620 ROMO.066, ROMO.320, ROMO.395, ROMO.410, Forest Alliance Vaccinium myrtillus Whortleberry Forest ROMO.514, ROMOAA.1093, ROMOAA.1114, Forest ROMOAA.5300, ROMOAA.66, ROMOAA.954, ROMOAA.955, ROMOAA.981, ROMOAA.982 Pseudotsuga menziesii Pseudotsuga menziesii / G4? CO:S3, Douglas-fir / Geyer's CEGL000430 ROMO.516 Forest Alliance Carex geyeri Forest ID:S4?, Sedge Forest MT:S4, OR:S3, WA:S1, WY

86 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Pseudotsuga menziesii Pseudotsuga menziesii / G3G4 CO:S3, WY Douglas-fir / Waxflower CEGL000438 ROMO.077, ROMO.409, ROMO.500, ROMO.622, Forest Alliance Jamesia americana Forest ROMOAA.1045, ROMOAA.370, ROMOAA.445, Forest ROMOAA.515, ROMOAA.516, ROMOAA.5227, ROMOAA.530, ROMOAA.5408, ROMOAA.5763, ROMOAA.5864, ROMOAA.5876, ROMOAA.630, ROMOAA.673, ROMOAA.881, ROMOAA.885, ROMOAA.965 Pseudotsuga menziesii Pseudotsuga menziesii / G4 CO:S1S2, Douglas-fir / Common CEGL000439 ROMO.187, ROMO.200, ROMO.221, ROMO.618, Forest Alliance Juniperus communis ID:S3, Juniper Forest ROMO.619, ROMOAA.29, ROMOAA.290, Forest MT:S4, ROMOAA.293, ROMOAA.331, ROMOAA.416, WY:S3S4 ROMOAA.473, ROMOAA.5192, ROMOAA.5214, ROMOAA.5261, ROMOAA.527, ROMOAA.5271, ROMOAA.54, ROMOAA.5432, ROMOAA.5734, ROMOAA.631, ROMOAA.696, ROMOAA.709, ROMOAA.741, ROMOAA.758, ROMOAA.804, ROMOAA.870, ROMOAA.884, ROMOAA.933 Pseudotsuga menziesii Pseudotsuga menziesii / G3G4 CO, ID:S3, Douglas-fir / Spike CEGL000904 ROMO.844 Woodland Alliance Leucopoa kingii WY:S3S4 Fescue Woodland Woodland Pseudotsuga menziesii Pseudotsuga menziesii / G4 CO:S4, Douglas-fir / Mountain CEGL000449 ROMO.501, ROMO.525, ROMO.760, ROMO.811, Forest Alliance Physocarpus MT:S1S2, Ninebark Forest ROMOAA.297, ROMOAA.34, ROMOAA.346, monogynus Forest NM:S4, ROMOAA.5431, ROMOAA.8005 WY:S2S3 Purshia tridentata Purshia tridentata / G1G2 CO:S1S2 Bitterbrush / Fringed CEGL001055 ROMO.8013, ROMOAA.1067, ROMOAA.5168 Shrubland Alliance Artemisia frigida / Sagebrush / Needle-and- Hesperostipa comata Thread Shrubland Shrubland Purshia tridentata Purshia tridentata / G2 CO:S2 Bitterbrush / Mountain CEGL001057 ROMO.040, ROMO.108, ROMO.203, ROMO.515, Shrubland Alliance Muhlenbergia montana Muhly Shrubland ROMO.524, ROMO.620, ROMO.625, ROMO.829, Shrubland ROMO.836, ROMO.840, ROMOAA.1097, ROMOAA.1113, ROMOAA.16, ROMOAA.173, ROMOAA.198, ROMOAA.267, ROMOAA.407, ROMOAA.415, ROMOAA.417, ROMOAA.418, ROMOAA.486, ROMOAA.5119, ROMOAA.5133, ROMOAA.5137, ROMOAA.519, ROMOAA.606, ROMOAA.729, ROMOAA.730, ROMOAA.801, ROMOAA.827, ROMOAA.828, ROMOAA.932

87 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Rhus trilobata Shrub Rhus trilobata Rocky G2 CO:S2 Squawbush Rocky CEGL002910 ROMO.624 Herbaceous Alliance Mountain Shrub Mountain Shrub Herbaceous Vegetation Herbaceous Vegetation Ribes montigenum Ribes montigenum GU CO:SU Western Prickly CEGL001133 ROMO.365, ROMO.398, ROMO.712, Shrubland Alliance Shrubland Gooseberry Shrubland ROMOAA.567 Rubus idaeus ssp. Rubus idaeus Scree GU CO:SU Red Raspberry Scree CEGL001134 ROMO.185, ROMO.386, ROMO.393, ROMO.394, strigosus Shrubland Shrubland Shrubland ROMOAA.704 Alliance Salix arctica Dwarf- Salix arctica - Salix G2Q CO, NM:S2 Arctic Willow - Snow CEGL001432 ROMO.015, ROMO.295, ROMO.473, ROMO.603 shrubland Alliance nivalis Dwarf-shrubland Willow Dwarf-shrubland Salix arctica Dwarf- Salix arctica / Geum G4 CO:S4? Arctic Willow / Ross' CEGL001430 ROMO.022, ROMO.461, ROMO.722, ROMO.723, shrubland Alliance rossii Dwarf-shrubland Avens Dwarf-shrubland ROMOAA.165, ROMOAA.219, ROMOAA.5104, ROMOAA.5967, ROMOAA.5992, ROMOAA.670 Salix brachycarpa Salix brachycarpa / G4 CO:S4, Short-fruit Willow / CEGL001135 ROMO.044, ROMO.047, ROMO.143, ROMO.244, Seasonally Flooded Mesic Forbs Shrubland WY:S3 Mesic Forbs Shrubland ROMO.380, ROMO.383, ROMO.498, ROMO.611, Shrubland Alliance ROMO.634, ROMO.714, ROMOAA.201, ROMOAA.202, ROMOAA.481, ROMOAA.482, ROMOAA.483, ROMOAA.5096, ROMOAA.5097, ROMOAA.5107, ROMOAA.5986, ROMOAA.5987, ROMOAA.5995, ROMOAA.609, ROMOAA.646, ROMOAA.778 Salix drummondiana Salix drummondiana / G3 AB?, BC?, Drummond's Willow / CEGL002667 ROMO.312, ROMO.719, Temporarily Flooded Calamagrostis CO:S3, Bluejoint Shrubland ROMO.737ROMOAA.1166 Shrubland Alliance canadensis Shrubland ID:S2, MT, WA:S2? Salix drummondiana Salix drummondiana / G4 BC, CO, Drummond's Willow / CEGL002631 ROMO.513, ROMOAA.945 Seasonally Flooded Carex utriculata ID:S3, Beaked Sedge Shrubland Shrubland Alliance Shrubland MT:S4, UT?, WA:S3, WY Salix drummondiana Salix drummondiana / G4 AB, CO:S4, Drummond's Willow / CEGL001192 ROMO.710 Temporarily Flooded Mesic Forbs Shrubland MT, WY Mesic Forbs Shrubland Shrubland Alliance

88 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Salix geyeriana Salix geyeriana - Salix G3 CO:S3 Geyer's Willow - CEGL001247 ROMO.269, ROMO.608, ROMO.738, ROMO.826, Temporarily Flooded monticola / Mountain Willow / ROMOAA.249, ROMOAA.5079 Shrubland Alliance Calamagrostis Bluejoint Shrubland canadensis Shrubland Salix geyeriana Salix geyeriana - Salix G3 CO:S3 Geyer's Willow - CEGL001223 ROMO.396, ROMO.736, ROMO.830 Temporarily Flooded monticola / Mesic Forbs Mountain Willow / Shrubland Alliance Shrubland Mesic Forbs Shrubland Salix geyeriana Salix geyeriana / G5 CO:S3, Geyer's Willow / CEGL001205 ROMO.316, ROMO.321, ROMO.734 Seasonally Flooded Calamagrostis ID:S4, Bluejoint Shrubland Shrubland Alliance canadensis Shrubland MT:S4, OR, UT:S2?, WY Salix geyeriana Salix geyeriana / Carex G3 CO:S3, Geyer's Willow / CEGL001206 ROMO.116, ROMO.206 Seasonally Flooded aquatilis Shrubland ID:S3, MT?, Aquatic Sedge Shrubland Alliance UT:S3?, Shrubland WY? Salix geyeriana Salix geyeriana / Carex G5 CO:S3, Geyer's Willow / Beaked CEGL001207 ROMO.212, ROMOAA.246, ROMOAA.999 Seasonally Flooded utriculata Shrubland ID:S4, Sedge Shrubland Shrubland Alliance MT:S5, NV, OR:S2, UT:S2S3, WY Salix geyeriana Salix geyeriana / Mesic G3? CO:S3, ID, Geyer's Willow / Mesic CEGL001210 ROMO.820, ROMO.821, ROMO.822, Temporarily Flooded Graminoids Shrubland NV, Graminoids Shrubland ROMOAA.5061, ROMOAA.5063, Shrubland Alliance UT:S2S3, ROMOAA.5066, ROMOAA.5071 WY Salix monticola Salix monticola / G3 CO:S3, NM? Mountain Willow / CEGL001222 ROMO.079, ROMO.742, ROMOAA.1030 Temporarily Flooded Calamagrostis Bluejoint Shrubland Shrubland Alliance canadensis Shrubland Salix monticola Salix monticola / Carex G3 CO:S3 Mountain Willow / CEGL002656 ROMO.287, ROMO.368, ROMO.606, Temporarily Flooded aquatilis Shrubland Aquatic Sedge ROMOAA.1125, ROMOAA.1126, ROMOAA.998 Shrubland Alliance Shrubland Salix monticola Salix monticola / Carex G3 CO:S3 Mountain Willow / CEGL002657 ROMO.261, ROMO.802, ROMO.819, Temporarily Flooded utriculata Shrubland Beaked Sedge Shrubland ROMOAA.8004 Shrubland Alliance

89 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Salix monticola Salix monticola / Mesic G3 CO:S3, Mountain Willow / CEGL002659 ROMO.211, ROMOAA.5180 Temporarily Flooded Graminoids Shrubland NM?, UT?, Mesic Graminoids Shrubland Alliance WY Shrubland Salix (reticulata, Salix nivalis / Geum GNR CO Snow Willow / Ross' CEGL005936 ROMO.062, ROMO.181, ROMOAA.5085 nivalis) Dwarf- rossii Dwarf-shrubland Avens Dwarf-shrubland shrubland Alliance Salix planifolia Salix planifolia / G4 CO:S3, Planeleaf Willow / CEGL001225 ROMO.054, ROMO.073, ROMO.160, ROMO.282, Temporarily Flooded Calamagrostis UT:S2?, Bluejoint Shrubland ROMO.354, ROMO.420, ROMOAA.1079, Shrubland Alliance canadensis Shrubland WY:S2? ROMOAA.1084, ROMOAA.5065, ROMOAA.5068, ROMOAA.5072, ROMOAA.5074, ROMOAA.5075, ROMOAA.5076, ROMOAA.5084, ROMOAA.568 Salix planifolia Salix planifolia / Caltha G4 CO:S4, WY Planeleaf Willow / CEGL002665 ROMO.057, ROMO.161, ROMO.275, ROMO.285, Seasonally Flooded leptosepala Shrubland White Marsh-marigold ROMO.496, ROMOAA.164, ROMOAA.485, Shrubland Alliance Shrubland ROMOAA.5087 Salix planifolia Salix planifolia / Carex G5 CO:S4, Planeleaf Willow / CEGL001227 ROMO.001, ROMO.026, ROMO.071, ROMO.136, Seasonally Flooded aquatilis Shrubland ID:S4, Aquatic Sedge ROMO.138, ROMO.213, ROMO.214, ROMO.307, Shrubland Alliance MT:S3, Shrubland ROMO.328, ROMO.440, ROMOAA.1140, NM:S4, ROMOAA.1145, ROMOAA.5054, ROMOAA.5058, ROMOAA.5062, UT:S2S3, ROMOAA.5092, ROMOAA.5093, WY ROMOAA.5947, ROMOAA.616, ROMOAA.764 Salix planifolia Salix planifolia / Carex G4 BC?, Planeleaf Willow / CEGL001229 ROMO.168, ROMO.249, ROMO.432, ROMO.470, Seasonally Flooded scopulorum Shrubland CO:S4?, Holm's Rocky Mountain ROMO.483, ROMOAA.1137, ROMOAA.1143, Shrubland Alliance ID:S3, Sedge Shrubland ROMOAA.261, ROMOAA.5051 WA:S3?, WY Salix planifolia Salix planifolia / Carex GNR CO, MT, Planeleaf Willow / CEGL005937 ROMO.228, ROMO.376, ROMO.382, Seasonally Flooded utriculata Shrubland UT, WY Beaked Sedge Shrubland ROMOAA.5053, ROMOAA.5073, Shrubland Alliance ROMOAA.5081, ROMOAA.5933

90 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Salix planifolia Salix planifolia / G2G3 CO, Planeleaf Willow / CEGL001230 ROMO.017, ROMO.031, ROMO.051, ROMO.059, Temporarily Flooded Deschampsia caespitosa UT:S2S3, Tufted Hairgrass ROMO.162, ROMO.271, ROMO.281, ROMO.286, Shrubland Alliance Shrubland WY Shrubland ROMO.374, ROMO.445, ROMOAA.238, ROMOAA.245, ROMOAA.257, ROMOAA.260, ROMOAA.5033, ROMOAA.5034, ROMOAA.5050, ROMOAA.5057, ROMOAA.5108, ROMOAA.597, ROMOAA.604 Salix planifolia Salix planifolia / Mesic G4 CO:S4, WY Planeleaf Willow / CEGL002893 ROMO.146, ROMO.334, ROMOAA.253, Seasonally Flooded Forbs Shrubland Mesic Forbs Shrubland ROMOAA.545 Shrubland Alliance Salix wolfii Seasonally Salix wolfii / Carex G4 CO:S3, Wolf Willow / Beaked CEGL001237 ROMO.735 Flooded Shrubland utriculata Shrubland ID:S4, Sedge Shrubland Alliance UT:S2?, WY:S2S3 Salix wolfii Salix wolfii / G3 CO, ID:S2, Wolf Willow / Tufted CEGL001238 ROMO.195, ROMO.274, ROMO.279, ROMO.280, Temporarily Flooded Deschampsia caespitosa MT:S3, Hairgrass Shrubland ROMO.739, ROMOAA.248 Shrubland Alliance Shrubland UT:S2S3, WY:S2S3 Salix wolfii Salix wolfii / Fragaria G4? CO, ID:S3, Wolf Willow / Virginia CEGL001239 ROMO.159, ROMO.273 Temporarily Flooded virginiana Shrubland WY:S3S4 Strawberry Shrubland Shrubland Alliance na Sedum lanceolatum ssp. na CO Lance-Leaf Stonecrop Local Type ROMO.448 lanceolatum Herbaceous Vegetation Herbaceous Vegetation Sibbaldia procumbens Sibbaldia procumbens - G3? CO:SU, Creeping Glow-wort - CEGL001933 ROMO.142 Herbaceous Alliance Polygonum bistortoides WY:S3? American Bistort Herbaceous Vegetation Herbaceous Vegetation Silene acaulis Silene acaulis G5? CO, MT:S5, Cushion Pink CEGL001934 ROMO.019, ROMO.259, ROMO.435, ROMO.454, Herbaceous Alliance Herbaceous Vegetation WY? Herbaceous Vegetation ROMO.455, ROMOAA.142, ROMOAA.5018

91 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

GLOBAL STATE ASSOCIATION ELEMENT NVC ALLIANCE NVC ASSOCIATION RANK RANK* COMMON NAME CODE PLOT CODES Sparganium Sparganium G4 CA, CO:SU, Greenfruit Bur-reed CEGL001990 ROMO.495, ROMO.602, ROMO.718, ROMO.727 angustifolium angustifolium ID, OR:S4, Herbaceous Vegetation Permanently Flooded Herbaceous Vegetation WA:S3S4 Herbaceous Alliance Sparse Nonvascular Sparse Nonvascular GNR CO Sparse Nonvascular CEGL002888 ROMO.230, ROMO.329, ROMO.389, ROMO.408, Vegetation (on rock Vegetation (on rock and Vegetation (on rock and ROMO.499, ROMOAA.1008, ROMOAA.1014, and unconsolidated unconsolidated unconsolidated ROMOAA.207, ROMOAA.209, ROMOAA.474, substrates) Alliance substrates) substrates) ROMOAA.5020, ROMOAA.5111, ROMOAA.5123, ROMOAA.5291, ROMOAA.5362, ROMOAA.5367, ROMOAA.5369, ROMOAA.5599, ROMOAA.570, ROMOAA.575, ROMOAA.576, ROMOAA.589, ROMOAA.590, ROMOAA.5965, ROMOAA.768 Trifolium Trifolium dasyphyllum G4 CO:S4?, Uinta Clover CEGL001935 ROMO.029, ROMO.351, ROMO.437, (dasyphyllum, nanum) Herbaceous Vegetation NM:S3S4 Herbaceous Vegetation ROMOAA.1138, ROMOAA.157, ROMOAA.167, Herbaceous Alliance ROMOAA.5008, ROMOAA.5030, ROMOAA.752 Trifolium Trifolium nanum GNR CO Tundra Clover CEGL005939 ROMO.178, ROMO.260, ROMO.337, ROMO.350, (dasyphyllum, nanum) Herbaceous Vegetation Herbaceous Vegetation ROMOAA.898 Herbaceous Alliance Trifolium parryi Trifolium parryi GU CO:SU, Parry's Clover CEGL001936 ROMO.258, ROMO.464, ROMO.471 Herbaceous Alliance Herbaceous Vegetation WY? Herbaceous Vegetation Vaccinium Vaccinium G4 AB, CO:S1?, (Dwarf Blueberry, CEGL001140 ROMO.058, ROMO.145, ROMO.243, ROMO.256, (caespitosum, (caespitosum, NV, WY Grouseberry) Dwarf- ROMO.262, ROMO.297, ROMO.335, ROMO.422, myrtillus, scoparium) scoparium) Dwarf- shrubland ROMO.434, ROMO.462, ROMOAA.1029, Dwarf-shrubland shrubland ROMOAA.1048, ROMOAA.1080, ROMOAA.163, ROMOAA.212, ROMOAA.235, ROMOAA.684, Alliance ROMOAA.812

92 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Photo-Interpretation and Map should be noted that some associations Units may occur in more than one map unit. We see this with the P. contorta / We recognized and delineated 46 map Juniperus communis Woodland which units on the true color aerial photographs we find at all elevations within P. for ROMO. Of these, 35 are vegetated contorta types. This is a many to one and received an accuracy assessment. relationship (many map units – one All map units were developed from a association). combination of an initial NVC vegetation classification provided by NatureServe with input from Park biologists and BOR ecologists, fieldwork, and preliminary photo- interpretation.

Appendix H details the descriptions and representative photos for all vegetation map units. Table 12 details each of the map units and salient associations.

A few map units have a one to one relationship with the vegetation associations, but most have several associations as part of each map unit. This is largely due to the fact that many associations are defined by the understory vegetation which is often not visible from overhead. Some map units may be modeled given known or inferred distribution. An example of this is the Pinus contorta map units. There are three P. contorta map units. One of these, the P. contorta / Rock Woodland is a one to one relationship. There is only one association within this map unit. The high and low elevation P. contorta map units are separated by the known distribution of the understory. Plot data shows us that, typically, the three Vaccinium species occur as a prominent member of the understory above 9,500 ft. We therefore modeled this distribution with some success (see results – accuracy assessment). This is a one to many relationship (one map unit – many associations). In addition, it

93 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 12. Map units and vegetation associations within each map unit.

Notes: 1 – This association occurs in multiple map units 2 – This association is a minor type in this map unit 3 – Local types are not part of the NVC

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes 1 Herbaceous Upland – Alpine Artemisia arctica ssp. arctica Herbaceous Vegetation CEGL001848 Carex elynoides - Geum rossii Herbaceous Vegetation CEGL001853 Carex pyrenaica Herbaceous Vegetation CEGL001860 1 Carex rupestris - Geum rossii Herbaceous Vegetation CEGL001861 Carex rupestris - Trifolium dasyphyllum Herbaceous Vegetation CEGL001863 Carex siccata - Geum rossii Herbaceous Vegetation CEGL001808 Deschampsia caespitosa Herbaceous Vegetation CEGL001599 1 Dryas octopetala - Carex rupestris Dwarf-shrub Herbaceous Vegetation CEGL001892 Festuca brachyphylla - Geum rossii var. turbinatum Herbaceous Vegetation CEGL001895 Geum rossii - Trifolium spp. Herbaceous Vegetation CEGL001970 Juncus parryi / Sibbaldia procumbens Herbaceous Vegetation CEGL005871 Kobresia myosuroides - Carex rupestris var. drummondiana Herbaceous CEGL001907 Vegetation Kobresia myosuroides - Geum rossii Herbaceous Vegetation CEGL001908 Salix arctica - Salix nivalis Dwarf-shrubland CEGL001432 Salix arctica / Geum rossii Dwarf-shrubland CEGL001430 Salix nivalis / Geum rossii Dwarf-shrubland CEGL005936 Sedum lanceolatum ssp. lanceolatum Herbaceous Vegetation Local Type 3 Sibbaldia procumbens - Polygonum bistortoides Herbaceous Vegetation CEGL001933 Trifolium dasyphyllum Herbaceous Vegetation CEGL001935 1 2 Herbaceous Upland - Alpine Fellfield Geum rossii - Minuartia obtusiloba Herbaceous Vegetation CEGL001965 Paronychia pulvinata - Silene acaulis Dwarf-shrubland CEGL001976 Silene acaulis Herbaceous Vegetation CEGL001934 Trifolium dasyphyllum Herbaceous Vegetation CEGL001935 1

94 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes Trifolium nanum Herbaceous Vegetation CEGL005939 Trifolium parryi Herbaceous Vegetation CEGL001936 4 Herbaceous Upland - Montane Bouteloua gracilis Herbaceous Vegetation CEGL001760 Bromus inermis - (Pascopyrum smithii) Semi-natural Herbaceous Vegetation CEGL005264 Danthonia parryi Herbaceous Vegetation CEGL001795 Elymus trachycaulus Alliance [Placeholder] Local Type 3 Festuca thurberi Subalpine Grassland Herbaceous Vegetation CEGL001631 Muhlenbergia montana - Hesperostipa comata Herbaceous Vegetation CEGL001647 Muhlenbergia montana Herbaceous Vegetation CEGL001646 5 Herbaceous Wetland - Cross Zone - Marsh Nuphar lutea ssp. polysepala Herbaceous Vegetation CEGL002001 Sparganium angustifolium Herbaceous Vegetation CEGL001990 6 Herbaceous Wetland - Cross Zone - Wetland Calamagrostis canadensis Western Herbaceous Vegetation CEGL001559 Caltha leptosepala - Rhodiola rhodantha Herbaceous Vegetation CEGL001957 Caltha leptosepala Herbaceous Vegetation CEGL001954 Cardamine cordifolia - Mertensia ciliata - Senecio triangularis Herbaceous CEGL002662 Vegetation Carex aquatilis - Carex utriculata Herbaceous Vegetation CEGL001803 Carex aquatilis Herbaceous Vegetation CEGL001802 Carex limosa Herbaceous Vegetation CEGL001811 Carex microptera Herbaceous Vegetation CEGL001792 Carex scopulorum - Caltha leptosepala Herbaceous Vegetation CEGL001823 Carex utriculata Herbaceous Vegetation CEGL001562 Eleocharis quinqueflora Herbaceous Vegetation CEGL001836 Glyceria grandis Herbaceous Vegetation CEGL003429 Juncus balticus Herbaceous Vegetation CEGL001838 Herbaceous Wetland - Subalpine / Alpine - 7 Carex illota Herbaceous Vegetation CEGL001876 Alpine Meadow Carex pyrenaica Herbaceous Vegetation CEGL001860 Danthonia intermedia Herbaceous Vegetation CEGL001794 Dasiphora fruticosa ssp. floribunda / Deschampsia caespitosa Shrubland CEGL001107 Dasiphora fruticosa ssp. floribunda Subalpine Shrubland CEGL003499

95 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes Deschampsia caespitosa - Caltha leptosepala Herbaceous Vegetation CEGL001882 Deschampsia caespitosa - Geum rossii Herbaceous Vegetation CEGL001884 Deschampsia caespitosa Herbaceous Vegetation CEGL001599 1 Geum rossii - Polygonum bistortoides Herbaceous Vegetation CEGL001967 Geum rossii - Sibbaldia procumbens Herbaceous Vegetation CEGL001969 Geum rossii - Trifolium spp. Herbaceous Vegetation CEGL001970 1 Juncus drummondii - Carex spp. Herbaceous Vegetation CEGL001905 Poa cusickii / Sibbaldia procumbens Herbaceous Vegetation Local Type 3 10 Rock (Alpine - Upper Subalpine) Aquilegia caerulea - Cirsium scopulorum Scree Sparse Vegetation CEGL001938 1 Rubus idaeus Scree Shrubland CEGL001134 1 Sparse non-vascular vegetation (on rock and unconsolidated substrates) CEGL002888 1 11 Rock (Foothill - Lower Subalpine) Sparse non-vascular vegetation (on rock and unconsolidated substrates) CEGL002888 1 120 Shrub - Riparian - Cross Zone > 9600 ft Alnus incana / Equisetum arvense Shrubland CEGL001146 1,2 Betula nana / Mesic Forbs - Mesic Graminoids Shrubland CEGL002653 1 Salix brachycarpa / Mesic Forbs Shrubland CEGL001135 1 Salix drummondiana / Mesic Forbs Shrubland CEGL001192 Salix geyeriana - Salix monticola / Mesic Forbs Shrubland CEGL001223 1,2 Salix monticola / Carex aquatilis Shrubland CEGL002656 1,2 Salix monticola / Carex utriculata Shrubland CEGL002657 1,2 Salix planifolia / Calamagrostis canadensis Shrubland CEGL001225 1 Salix planifolia / Caltha leptosepala Shrubland CEGL002665 Salix planifolia / Carex aquatilis Shrubland CEGL001227 1 Salix planifolia / Carex scopulorum Shrubland CEGL001229 Salix planifolia / Carex utriculata Shrubland CEGL005937 1 Salix planifolia / Deschampsia caespitosa Shrubland CEGL001230 Salix planifolia / Mesic Forbs Shrubland CEGL002893 1 Salix planifolia / Mesic Forbs Shrubland CEGL002893 Salix wolfii / Carex utriculata Shrubland CEGL001237 Salix wolfii / Deschampsia caespitosa Shrubland CEGL001238 Salix wolfii / Fragaria virginiana Shrubland CEGL001239

96 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes 121 Shrub - Riparian - Cross Zone < 9600 ft Acer glabrum Drainage Bottom Shrubland CEGL001062 Alnus incana - Salix (monticola, lucida, ligulifolia) Shrubland CEGL002651 Alnus incana - Salix drummondiana Shrubland CEGL002652 Alnus incana / Equisetum arvense Shrubland CEGL001146 1 Alnus incana / Mesic Graminoids Shrubland CEGL001148 Betula nana / Mesic Forbs - Mesic Graminoids Shrubland CEGL002653 1 Betula occidentalis / Mesic Graminoids Shrubland CEGL002654 Dasiphora fruticosa ssp. floribunda Shrubland [Provisional] CEGL001105 Salix drummondiana / Calamagrostis canadensis Shrubland CEGL002667 Salix drummondiana / Carex utriculata Shrubland CEGL002631 Salix geyeriana - Salix monticola / Calamagrostis canadensis Shrubland CEGL001247 Salix geyeriana - Salix monticola / Mesic Forbs Shrubland CEGL001223 1 Salix geyeriana / Calamagrostis canadensis Shrubland CEGL001205 Salix geyeriana / Carex aquatilis Shrubland CEGL001206 Salix geyeriana / Carex utriculata Shrubland CEGL001207 Salix geyeriana / Mesic Graminoids Shrubland CEGL001210 Salix monticola / Calamagrostis canadensis Shrubland CEGL001222 1 Salix monticola / Carex aquatilis Shrubland CEGL002656 1 Salix monticola / Carex utriculata Shrubland CEGL002657 1 Salix monticola / Mesic Graminoids Shrubland CEGL002659 Salix planifolia / Calamagrostis canadensis Shrubland CEGL001225 Salix planifolia / Carex aquatilis Shrubland CEGL001227 1 Salix planifolia / Carex utriculata Shrubland CEGL005937 1 13 Shrub Upland - Alpine Betula nana - Salix brachycarpa Shrubland CEGL005828 Ribes montigenum Shrubland CEGL001133 Salix brachycarpa / Mesic Forbs Shrubland CEGL001135 1 Salix planifolia / Mesic Forbs Shrubland CEGL002893 1 Vaccinium (caespitosum, scoparium) Dwarf-shrubland CEGL001140 14 Shrub Upland - Lower Montane Cercocarpus montanus / Hesperostipa comata Shrubland CEGL001092 Cercocarpus montanus / Muhlenbergia montana Shrubland CEGL002914

97 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes Rhus trilobata Rocky Mountain Shrub Herbaceous Vegetation CEGL002910 Shrub Upland - Lower Montane - Big Artemisia tridentata ssp. vaseyana - (Purshia tridentata) / Muhlenbergia 141 CEGL005827 Sagebrush montana - (Hesperostipa comata ssp. comata) Shrubland 142 Shrub Upland - Lower Montane - Bitterbrush Purshia tridentata / Artemisia frigida / Hesperostipa comata Shrubland CEGL001055 Purshia tridentata / Muhlenbergia montana Shrubland CEGL001057 15 Riparian Aspen Populus tremuloides - Picea pungens Forest CEGL000535 Populus tremuloides / Acer glabrum Forest CEGL000563 Populus tremuloides / Alnus incana Forest CEGL001150 Populus tremuloides / Calamagrostis canadensis Forest CEGL000574 Populus tremuloides / Phleum pratense Semi-natural Forest CEGL005829 18 Upland Aspen Populus balsamifera / Pteridium aquilinum Woodland Local Type 3 Populus tremuloides / Juniperus communis Forest CEGL000587 Populus tremuloides / Physocarpus monogynus Forest CEGL005932 Populus tremuloides / Poa pratensis Forest CEGL003148 Populus tremuloides / Prunus virginiana Forest CEGL000596 Populus tremuloides / Pteridium aquilinum Forest CEGL000597 Populus tremuloides / Shepherdia canadensis Forest CEGL000606 Populus tremuloides / Thalictrum fendleri Forest CEGL000619 Populus tremuloides / Vaccinium myrtillus Forest CEGL000620 161 Mixed conifer with aspen (Ponderosa Pine) Populus tremuloides - Pinus ponderosa Rocky Mountain Forest CEGL000541 162 Mixed conifer with aspen (Lodgepole Pine) Populus tremuloides - Pinus contorta / Juniperus communis Forest CEGL000537 163 Mixed conifer with aspen (Douglas-fir) Populus tremuloides - Pseudotsuga menziesii / Juniperus communis Forest CEGL000545 164 Mixed conifer with aspen (Spruce - Fir) Populus tremuloides - Abies lasiocarpa / Juniperus communis Forest CEGL000527 Upper Montane - Mixed Conifer - Riparian > 190 Abies lasiocarpa - Picea engelmannii / Acer glabrum Forest CEGL000294 9600 ft Abies lasiocarpa - Picea engelmannii / Calamagrostis canadensis Forest CEGL000300 Abies lasiocarpa - Picea engelmannii / Mertensia ciliata Forest CEGL002663 Abies lasiocarpa - Picea engelmannii / Salix drummondiana Forest CEGL000327 1,2 Picea engelmannii / Calamagrostis canadensis Forest CEGL002678 1 Upper Montane - Mixed Conifer - Riparian < 191 Abies lasiocarpa - Picea engelmannii / Alnus incana Forest CEGL000296 9600 ft

98 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes Abies lasiocarpa - Picea engelmannii / Calamagrostis canadensis Forest CEGL000300 1 Abies lasiocarpa - Picea engelmannii / Salix drummondiana Forest CEGL000327 1 Picea engelmannii / Calamagrostis canadensis Forest CEGL002678 1 Picea engelmannii / Equisetum arvense Forest CEGL005927 20 Montane Douglas Fir Pseudotsuga menziesii / Carex geyeri Forest CEGL000430 Pseudotsuga menziesii / Jamesia americana Forest CEGL000438 Pseudotsuga menziesii / Juniperus communis Forest CEGL000439 Pseudotsuga menziesii / Leucopoa kingii Woodland CEGL000904 1 Pseudotsuga menziesii / Physocarpus monogynus Forest CEGL000449 22 Subalpine mixed conifer Abies lasiocarpa - Picea engelmannii / Juniperus communis Woodland CEGL000919 Abies lasiocarpa - Picea engelmannii / Moss Forest CEGL000321 Abies lasiocarpa - Picea engelmannii / Polemonium pulcherrimum Forest CEGL000373 Abies lasiocarpa - Picea engelmannii / Vaccinium caespitosum Forest CEGL000340 Abies lasiocarpa - Picea engelmannii / Vaccinium myrtillus Forest CEGL000343 Abies lasiocarpa - Picea engelmannii / Vaccinium scoparium Forest CEGL000344 Picea engelmannii / Moss Forest CEGL000371 Picea engelmannii / Trifolium dasyphyllum Forest CEGL000377 Picea engelmannii / Vaccinium myrtillus Forest CEGL000379 Picea engelmannii / Vaccinium scoparium Forest CEGL000381 23 Lodgepole - high elevation > 9500 ft. Pinus contorta / Juniperus communis Woodland CEGL000764 1 Pinus contorta / Vaccinium caespitosum Forest CEGL000168 Pinus contorta / Vaccinium myrtillus Forest CEGL005935 1 Pinus contorta / Vaccinium scoparium Forest CEGL000172 24 Lodgepole - low elevation < 9500 ft. Pinus contorta / Arctostaphylos uva-ursi Forest CEGL000134 Pinus contorta / Carex geyeri Forest CEGL000141 Pinus contorta / Carex rossii Forest CEGL000144 Pinus contorta / Jamesia americana Forest CEGL005933 Pinus contorta / Juniperus communis Woodland CEGL000764 1 Pinus contorta / Shepherdia canadensis Forest CEGL000163 Pinus contorta / Vaccinium myrtillus Forest CEGL005935 1,2

99 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes 26 Lodgepole pine - Rock Pinus contorta / Rock Woodland CEGL005934 1 32 Cottonwood Populus angustifolia / Betula occidentalis Woodland CEGL000648 Populus angustifolia / Bromus inermis Semi-natural Woodland CEGL005847 33 Juniper Woodland Juniperus scopulorum / Purshia tridentata Woodland CEGL000749 34 Ponderosa Pine - Graminoid Pinus ponderosa / Carex geyeri Woodland CEGL000182 Pinus ponderosa / Carex inops ssp. heliophila Woodland CEGL000849 Pinus ponderosa / Carex rossii Forest CEGL000183 Pinus ponderosa / Leucopoa kingii Woodland CEGL000186 Pinus ponderosa / Muhlenbergia montana Woodland CEGL000862 Pseudotsuga menziesii / Leucopoa kingii Woodland CEGL000904 1,2 35 Ponderosa Pine - Rockland Pinus ponderosa / Rockland Woodland CEGL000877 36 Ponderosa Pine - Shrubland Pinus ponderosa / Arctostaphylos uva-ursi Woodland CEGL000844 Pinus ponderosa / Artemisia tridentata ssp. vaseyana Woodland CEGL002794 Pinus ponderosa / Cercocarpus montanus Woodland CEGL000851 Pinus ponderosa / Juniperus communis Woodland CEGL000859 Pinus ponderosa / Physocarpus monogynus Forest CEGL000190 Pinus ponderosa / Purshia tridentata Woodland CEGL000867 Pinus ponderosa / Ribes cereum Forest CEGL000199 38 Limber Pine Pinus flexilis / Arctostaphylos uva-ursi Woodland CEGL000802 Pinus flexilis / Juniperus communis Woodland CEGL000807 41 Disturbance – Dead and Down 43 Blue Spruce Picea pungens Temporarily Flooded Woodland Alliance A.567 Picea pungens / Alnus incana Woodland CEGL000894 39 Ribbon Forests (subset of map unit 400) Abies lasiocarpa – Picea engelmannii / Salix (brachycarpa, glauca) 400 Krummholz CEGL000986 Krummholz Shrubland Abies lasiocarpa - Picea engelmannii Krummholz Shrubland CEGL000985 Pinus flexilis Krummholz Shrubland Local Type 3 9 Alpine - Ice Field - Glacier 46 Talus Aquilegia caerulea - Cirsium scopulorum Scree Sparse Vegetation CEGL001938 1

100 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Veg Code MAP UNIT NAME ASSOCIATION Elcode Notes Rubus idaeus Scree Shrubland CEGL001134 1 Sparse non-vascular vegetation (on rock and unconsolidated substrates) CEGL002888 1 48 Exposed Soil - Man made 49 Cliff Face - Bare Soil Sparse non-vascular vegetation (on rock and unconsolidated substrates) CEGL002888 51 Streams - Rivers 52 Natural Lakes - Ponds 53 Reservoir - Stock Tanks 999 Un-vegetated Surfaces

101 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

system codes. Using these items one can Vegetation Map then link to external databases that may contain more information than provided A total of 427,438 acres (172,982 ha) here. For example, if one links the El Code comprising ROMO and its environs was in Table 21 to the El Code in the GIS mapped. The area mapped within the Park database one can query for a range of G or S boundary was 289,469 acres (117,147 ha). ranked Ecological Systems, the plots in each Forty-six map units were used to describe or, after further linking to the plots database, the landscape. Of these, twelve were un- a query for specific species with specific vegetated map units. Of all the map units, cover values, heights or any other item the most frequently occurring within the recorded as part of this effort. entire mapping area was Map Unit 22 Spruce – Fir with 4,521 polygons ranging in The utility of this map extends from the very size from 0.03 acres (0.01 ha) to 1,400 acres basic to the very involved. Much depends (567 ha). The most abundant map unit in upon the sophistication and imagination of terms of area was also Map Unit 22, Spruce the user or the investigator. For the more – Fir, covering 103,841 acres or about 24% advanced investigations one might expect of the project area. Spatial statistics for each the need for a GIS analyst. of the map units are listed in Table 13. A reduced statistics table is included for areal coverage of map units within the Park (Table 14). Polygon size ranged from 0.02 acres (0.01 ha) to 3,254 acres (1,317 ha) with the mean polygon size being 12 acres (4.8 ha).

The individual map unit statistics are important in that they reveal so much more than just the mean size. Often the mean area for each map unit is highly skewed. For example, many small polygons will show a higher frequency for the small polygons yet a few large polygons may represent the greatest area. The use of mean as a summary statistic may then be highly misleading.

Particularly useful is the several vegetation codes that describe each polygon. We have the project specific vegetation code developed specifically for this mapping effort that has considerable local detail. We have also cross-walked the project specific vegetation code to several others that will allow for analysis at various other scales and perspectives. These include two Anderson type landcover codes and the ecological

102 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 13. Spatial statistics for all map units in the mapping area.

Min Max Ave SD Min Max Ave Sum SD Var Min Max Ave Sum SD Var VEG CODE Map Unit Name Count ELEV ELEV ELEV ELEV ACRES ACRES ACRES ACRES ACRES ACRES HECT HECT HECT HECT HECT HECT (Ft) (Ft) (Ft) (Ft) 1 Herbaceous Upland Alpine > 1243 9603 13146 11196 649 9600 ft 0.1 447.5 11.6 14407.3 33.1 1093.5 0.0 181.1 4.7 5830.6 13.4 179.1 2 Herbaceous Upland Alpine 9504 13281 11455 597 Fellfield 0.1 1682.5 28.8 19712.6 113.3 12829.6 0.0 680.9 11.7 7977.6 45.8 2101.2 4 Herbaceous Upland Montane 1519 7005 9593 8402 501 < 9600 ft 0.0 385.5 6.1 9201.6 19.6 382.4 0.0 156.0 2.5 3723.9 7.9 62.6 5 Herbaceous Wetland Cross 59 7444 11197 9326 999 Zone - Marsh 0.1 9.3 1.3 76.2 1.7 2.8 0.0 3.8 0.5 30.8 0.7 0.5 6 Herbaceous Wetland Cross 1935 7356 12395 10227 1013 Zone - Wetland 0.0 273.0 3.8 7355.9 12.3 150.1 0.0 110.5 1.5 2976.9 5.0 24.6 7 Herbaceous Wetland Subalpine / Alpine - Alpine 1374 8376 13333 11182 667 Mead 0.0 928.8 15.6 21454.5 58.5 3424.9 0.0 375.9 6.3 8682.5 23.7 560.9 9 Glacier 93 0.1 33.0 3.9 358.5 5.9 34.3 0.0 13.3 1.6 145.1 2.4 5.6 9616 13317 11866 609 10 Rock (Alpine-Upper 559 8120 13615 11234 848 Subalpine) 0.0 274.7 6.0 3366.6 19.3 373.4 0.0 111.2 2.4 1362.4 0.3 61.2 11 Rock (Foothill-Lower 2229 7159 11525 9380 879 Subalpine) 0.0 25.9 1.3 2893.2 2.1 4.4 0.0 10.5 0.5 1170.9 0.9 0.7 13 Shrub Upland Alpine 729 0.1 443.2 6.3 4622.7 22.8 521.6 0.0 179.4 2.6 1870.8 9.2 85.4 8737 12401 11050 603 14 Shrub Upland Lower Montane 490 7034 10676 8376 646 - Undifferentiated 0.2 166.8 6.7 3300.4 14.7 216.6 0.1 67.5 2.7 1335.6 6.0 35.5 15 Riparian Aspen 324 0.1 53.7 2.9 934.5 5.3 27.8 0.0 21.7 1.2 378.2 2.1 4.6 7001 10102 8483 703 18 Upper Montane Aspen 1349 0.1 62.1 1.9 2505.8 3.8 14.7 0.0 25.1 0.8 1014.1 1.6 2.4 7211 10955 8904 623 20 Lower Montane Douglas-fir 2282 0.1 402.7 16.7 38148.7 29.3 858.6 0.0 163.0 6.8 15438.6 11.9 140.6 7083 11312 8581 637 22 Subalpine Mixed Conifer 4521 0.0 1400.9 23.0 103841.6 76.7 5876.3 0.0 566.9 9.3 42024.1 31.0 962.4 8054 11965 10435 696 23 Lodgepole Pine - High 1039 9501 11256 10067 408 Elevation > 9500 ft 0.1 668.3 25.5 26498.6 52.9 2801.3 0.0 270.5 10.3 10723.8 21.4 458.8 24 Lodgepole Pine - Low 2441 7333 9498 8831 370 Elevation < 9500 ft 0.1 904.1 18.0 43933.6 46.6 2174.6 0.0 365.9 7.3 17779.7 18.9 356.1 26 Lodgepole Pine - Rock 145 0.3 181.4 8.3 1198.7 18.0 324.1 0.1 73.4 3.4 485.1 7.3 53.1 7949 10955 9456 688 32 Cottonwood 29 0.1 11.7 2.7 77.5 3.0 8.8 0.0 4.7 1.1 31.4 1.2 1.4 7014 9698 8052 718 33 Juniper 103 0.2 53.1 8.6 890.0 10.1 101.8 0.1 21.5 3.5 360.2 4.1 16.7 7067 10059 8364 620 34 Ponderosa pine Graminoid 1137 0.0 487.3 12.2 13863.7 24.3 589.6 0.0 197.2 4.9 5610.6 9.8 96.6 7129 9964 8291 463 35 Ponderosa pine Rockland 805 0.2 350.8 14.5 11692.8 30.6 936.5 0.1 142.0 5.9 4732.0 12.4 153.4 7254 10453 8345 501 36 Ponderosa pine Shrubland 797 0.2 208.0 15.7 12530.5 23.9 573.3 0.1 84.2 6.4 5071.0 9.7 93.9 7011 10535 8184 467 38 Subalpine Limber Pine 509 0.1 241.3 12.9 6552.0 22.6 509.4 0.1 97.6 5.2 2651.6 9.1 83.4 8287 11319 10187 605 39 Ribbon Forest - Islands 14 0.3 20.3 7.9 110.7 7.3 53.0 0.1 8.2 3.2 44.8 3.0 8.7 11132 11571 11389 132 41 Disturbance - Dead and Down 108 0.2 299.0 10.3 1111.0 33.6 1126.9 0.1 121.0 4.2 449.6 13.6 184.6 7815 11256 9407 852 43 Blue Spruce 95 0.1 43.0 5.2 495.1 7.2 51.1 0.0 17.4 2.1 200.4 2.9 8.4 7083 10341 8318 604 46 Talus 1561 0.1 2361.7 10.8 16847.4 69.1 4770.0 0.0 955.8 4.4 6818.1 28.0 781.2 7746 14006 10624 760 47 Outwash 5 1.0 110.3 26.5 132.3 47.3 2233.2 0.4 44.6 10.7 53.5 19.1 365.8 8569 11867 10176 1426 48 Exposed Soil Man-made 50 0.1 142.5 10.4 521.6 27.8 772.4 0.1 57.7 4.2 211.1 11.3 126.5 8054 11663 8781 759 49 Cliff face – Bare soil / Rock 162 0.2 3255.0 103.4 16754.6 307.9 94778.9 0.1 1317.3 41.9 6780.5 124.6 15522.7 7067 13510 11282 1162 51 Streams - rivers 38 0.1 20.1 5.2 199.0 5.4 29.5 0.0 8.2 2.1 80.5 2.2 4.8 7096 11417 8731 921 52 Natural Lakes - Ponds 531 0.0 829.4 4.6 2435.6 37.5 1406.6 0.0 355.7 1.9 985.7 15.2 230.4 7093 13192 10304 1242

103 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Min Max Ave SD Min Max Ave Sum SD Var Min Max Ave Sum SD Var VEG CODE Map Unit Name Count ELEV ELEV ELEV ELEV ACRES ACRES ACRES ACRES ACRES ACRES HECT HECT HECT HECT HECT HECT (Ft) (Ft) (Ft) (Ft) 53 Reservoirs – Stock Tanks 51 0.1 1299.8 40.8 2081.5 193.0 37242.5 0.1 526.0 16.5 842.4 78.1 6099.5 7474 10613 8488 616 Shrub Riparian Cross Zone > 120 469 9609 12113 10850 588 9600 ft 0.1 562.7 8.3 3866.9 35.0 1224.0 0.0 227.7 3.3 1564.9 14.2 200.5 Shrub Riparian Cross Zone < 121 506 7126 11023 8548 520 9600 ft 0.1 291.4 6.5 3284.0 20.6 424.6 0.0 117.9 2.6 1329.0 8.3 69.5 Shrub Upland Lower Montane 141 156 7274 9527 8491 313 - Big Sagebrush 0.2 205.8 6.8 1059.7 19.0 361.6 0.1 83.3 2.8 428.8 7.7 59.2 Shrub Upland Lower Montane 142 63 7050 9239 8302 535 - Bitterbrush 0.4 28.3 3.4 215.8 4.2 17.6 0.2 11.5 1.4 87.3 1.7 2.9 Mixed Conifer with Aspen 161 403 7044 9796 8462 448 (Ponderosa pine) 0.1 70.3 5.1 2065.5 7.9 63.1 0.0 28.5 2.1 835.9 3.2 10.3 Mixed Conifer with Aspen 162 461 7851 10702 9006 468 (Lodgepole Pine) 0.1 104.3 4.0 1832.8 8.9 78.7 0.0 42.2 1.6 741.7 3.6 12.9 Mixed Conifer with Aspen 163 532 7451 10371 8810 494 (Douglas-fir) 0.1 122.0 5.3 2811.8 11.5 132.7 0.0 49.4 2.1 1137.9 4.7 21.7 Mixed Conifer with Aspen 164 253 8107 10869 9327 533 (Spruce - Fir) 0.1 65.0 5.1 1292.6 8.2 67.8 0.0 26.3 2.1 523.1 3.3 11.1 Riparian Upper Montane 190 662 8517 11699 10143 927 Mixed Conifer > 8500 ft 0.1 372.9 12.5 8225.4 30.3 915.3 0.0 150.9 5.1 3340.9 12.2 149.9 Riparian Lower Montane 191 134 7228 8497 8083 299 Mixed Conifer < 8500 ft 0.5 35.2 7.2 960.4 7.2 52.5 0.2 14.2 2.9 388.7 2.9 8.6 400 Krummholz 3358 0.0 94.9 2.7 9121.1 6.6 43.1 0.0 38.4 1.1 3691.3 2.7 7.1 10138 12126 11325 282 999 Un-vegetated surfa 144 0.3 721.7 18.4 2645.5 64.9 4208.9 0.1 292.1 704 1070.6 26.3 689.3 7431 11824 8403 829

104 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 14. Summary area statistics for map units within ROMO.

VEG CODE COUNT Sum ACRES Sum HECTARES 1 1052 13503 5464 2 532 18528 7498 4 396 1939 785 5 25 38 15 6 1518 5759 2331 7 1058 19200 7770 9 74 260 105 10 452 3052 1235 11 1201 1720 696 13 643 4135 1673 14 186 1334 540 15 107 296 120 18 512 1123 454 20 741 10887 4406 22 3547 85736 34697 23 852 22456 9088 24 1232 23852 9653 26 107 952 385 32 17 31 12 33 50 359 145 34 196 1391 563 35 247 3007 1217 36 205 3586 1451 38 401 5392 2182 39 14 111 45 41 89 899 364 43 22 115 46 46 1284 14768 5976 47 5 132 54 48 20 249 101 49 130 15933 6448 51 18 124 50 52 410 2073 839 53 7 1324 536 120 393 3522 1425 121 284 2575 1042 141 46 274 111 142 31 128 52 161 90 499 202 162 212 837 339 163 187 999 404 164 137 629 255 190 535 6721 2720 191 39 199 80 400 2747 7732 3129 999 28 1174 475

105 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Fuzzy Accuracy Assessment

The use of “fuzzy” techniques to considered “correct. The overall map describe the accuracy of thematic maps accuracies for each of the fuzzy classes is a useful if somewhat ambiguous tool. are outlined in Table 15 and include a Now one is forced to interpret the 90% confidence interval and Kappa thematic accuracy of a product from statistic. The contingency table detailed multiple perspectives and a number of results are shown in Table 16, Table 17 caveats. There is no “one” figure to use and Table 18 for fuzzy classes 5, 4 and 3 as an estimate for either overall or respectively. Each map unit is analyzed individual map unit accuracies. It is now in terms of its individual accuracy for standard to couch the results in statistical omission and commission (producer’s parlance of confidence intervals and error and user’s error respectively) for sample sizes. Its use in many thematic three levels of fuzzy accuracy and products today originates from the includes a 90% confidence interval. A recognition that the binary approach of summary table of all map unit accuracies either “right” or “wrong” belies the true for both omission and commission for all nature of most map units and even the fuzzy levels is shown in Table 19. Mean view from the person or persons errors of omission and commission for providing the “reference” data. each fuzzy level are shown in the last row in Table 19 and separately in Table The great utility of a fuzzy approach is 20. Individual map unit metrics are the acknowledgement of degrees of discussed below. correctness. Only occasionally do map units have discrete boundaries; more Comparison of mean overall and often grading into one another over omission/commission accuracies distances ranging from a few to between fuzzy levels: Predictably, hundreds of meters. The necessity of overall map accuracy increases as one drawing discrete lines representing non- relaxes requirements for individual map discrete entities requires other than a unit membership. Table 15 shows the binary approach. increasing overall accuracy from fuzzy level 5 to fuzzy level 3. We also We performed a fuzzy accuracy include a Kappa statistic as a metric of assessment on the digital thematic map the overall accuracy. This statistic for ROMO. Only vegetated map units assumes that a certain number of correct were sampled. Table 10 describes the 5 classifications will occur by chance. fuzzy classes used during this analysis. Therefore, the Kappa statistic penalizes This concept and class descriptions was the overall map accuracy. The mean first described by Gopal and Woodcock error for omission and commission also (1994) using fuzzy set theory described increase as one relaxes map unit by Zadeh (1965). A fuzzy class was membership requirements. only analyzed using a contingency table for the top three fuzzy classes that are

106 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 15. Overall map accuracies for each fuzzy class.

Standard Overall Map Overall Map Deviation (90% - Fuzzy Class Accuracy Accuracy (Kappa) two tailed) 5 50.3 2.4 47.5 4 74.7 2.1 72.4 3 86.7 1.7 80.9

107 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park Commission Error Commission Fuzzy 5 +/ -(90%Conf. %Correct %Correct Interval) Interval)

Reference (Accuracy Assessment Field Data) Sum

1 2 4 5 6 7 13 14 15 18 20 22 23 24 26 32 33 34 35 36 38 41 43 120 121 141 142 161 162 163 164 190 191 400

1 31 8 0 0 0 13 3 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 59.0 52.5% 11.5% 2 3 7 0 0 0 4 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 17.0 41.2% 22.6% 4 0 0 28 0 0 0 0 1 0 0 3 0 0 0 0 0 0 1 0 1 0 0 0 0 0 1 2 0 0 0 0 0 0 0 37.0 75.7% 13.0%

5 0 0 1 7 7 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 18.0 38.9% 21.7% Users Accuracy (Commission Error) Conf 6 0 0 1 0 24 2 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 33.0 72.7% 14.3% 7 12 1 0 0 4 23 7 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 1 51.0 45.1% 12.4% 13 2 0 0 0 0 1 13 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 12 1 0 0 0 1 0 1 0 0 4 37.0 35.1% 14.3% 14 0 0 3 0 0 0 0 4 0 2 0 0 0 0 0 0 0 0 0 9 0 0 0 0 2 0 8 0 0 0 0 0 0 0 28.0 14.3% 12.7% 15 0 0 0 0 1 0 0 1 12 7 0 0 0 0 0 3 0 0 0 0 0 0 0 0 1 0 0 2 0 1 0 1 1 0 30.0 40.0% 16.4% 18 0 0 0 0 0 0 0 0 7 20 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 1 0 0 1 0 0 0 32.0 62.5% 15.6% 20 0 0 0 0 0 1 0 0 0 0 15 2 3 8 2 0 0 6 3 14 2 0 0 0 0 0 0 1 0 0 0 0 0 0 57.0 26.3% 10.5% 22 0 0 0 0 0 0 0 0 0 0 0 59 4 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 1 67.0 88.1% 7.3% 23 0 0 0 0 0 0 0 0 0 0 0 6 17 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 25.0 68.0% 17.3% 24 0 0 0 0 0 0 0 0 0 0 1 1 0 34 1 0 0 0 1 2 1 0 0 0 0 0 0 0 1 0 0 0 1 0 43.0 79.1% 11.4% 26 0 0 0 0 0 0 0 0 0 1 2 7 5 1 9 0 0 0 2 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 28.0 32.1% 16.3% 32 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4.0 25.0% 48.1% 33 1 0 0 0 0 0 0 2 0 0 3 0 0 0 0 0 13 2 1 7 0 0 0 0 0 0 1 0 0 0 0 0 0 0 30.0 43.3% 16.5% 34 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 14 0 12 0 0 0 0 0 0 1 1 0 1 0 0 0 0 31.0 45.2% 16.3% 35 0 0 0 0 0 0 0 0 0 0 7 0 1 2 0 0 0 2 10 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 31.0 32.3% 15.4% idence Interval is 36 0 0 0 0 0 0 0 1 0 0 2 0 0 1 0 0 0 7 1 19 0 0 0 0 0 1 0 0 0 0 0 0 0 0 32.0 59.4% 15.8% 38 0 1 0 0 0 0 0 0 0 0 3 10 3 2 1 0 0 0 0 0 12 0 0 0 0 1 0 0 0 0 0 0 0 0 34.0 35.3% 15.0%

Sample Data (Polygon Map Data) 41 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 3 1 2 1 47 0 0 0 0 0 0 0 0 1 0 0 1 57.0 82.5% 9.2% 43 0 0 0 0 0 0 0 0 0 0 2 6 0 1 0 0 0 1 0 2 0 0 7 0 0 0 0 0 0 0 1 1 1 0 22.0 31.8% 18.6% 120 2 0 0 0 6 6 3 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 22 0 0 0 0 0 0 0 0 0 0 41.0 53.7% 14.0% 121 0 0 0 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 27 0 0 0 0 1 0 0 0 1 32.0 84.4% 12.1% 90% two-sided limit two-sided 90% 141 0 0 2 0 1 0 0 7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 19 4 0 0 0 0 0 0 0 33.0 57.6% 15.7% 142 0 0 4 0 0 0 0 6 0 0 1 0 0 0 0 0 2 0 0 3 0 0 0 0 0 0 13 0 0 0 0 0 0 0 29.0 44.8% 16.9% 161 0 0 0 0 0 0 0 0 1 3 1 0 0 1 0 0 1 5 1 5 0 0 0 0 2 0 0 5 1 1 0 0 0 0 27.0 18.5% 14.1% 162 0 0 0 0 0 0 0 0 0 0 2 7 2 0 1 0 0 0 0 0 0 0 1 0 0 0 0 1 15 1 4 0 0 0 34.0 44.1% 15.5% 163 0 0 0 0 0 0 0 0 0 0 1 4 0 4 1 0 0 0 0 2 1 0 0 0 0 0 0 2 4 3 2 1 0 0 25.0 12.0% 12.7% 164 0 0 0 0 0 0 0 0 1 3 2 4 2 2 0 0 0 0 0 1 0 0 0 0 0 0 0 0 4 1 7 0 3 0 30.0 23.3% 14.4% 190 0 0 0 0 0 0 0 0 0 0 0 28 2 4 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 1 6 0 1 44.0 13.6% 9.6% 191 0 0 0 0 0 0 0 0 0 0 3 1 0 0 1 0 0 2 0 6 0 0 4 0 0 0 0 1 1 0 0 0 4 0 23.0 17.4% 15.2% 400 0 0 0 0 0 2 0 0 0 0 0 10 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 47 61.0 77.0% 9.7%

Sum 51 17 39 7 44 52 30 23 22 40 50 149 39 63 16 4 16 43 20 98 22 48 12 40 36 23 29 14 30 9 19 10 10 57 % Accurate 60.8% 41.2% 71.8% 100.0% 54.5% 44.2% 43.3% 17.4% 54.5% 50.0% 30.0% 39.6% 43.6% 54.0% 56.3% 25.0% 81.3% 32.6% 50.0% 19.4% 54.5% 97.9% 58.3% 55.0% 75.0% 82.6% 44.8% 35.7% 50.0% 33.3% 36.8% 60.0% 40.0% 82.5% Total Correct = 594 +/ - (90% Total Samples = 1182 Conf. 12.2% 22.6% 13.1% 7.1% 7.1% 12.3% 16.5% 15.2% 19.7% 14.3% 11.7% 6.9% 14.3% 11.1% 23.5% 48.1% 19.2% 12.9% 20.9% 7.1% 19.7% 4.4% 27.6% 14.2% 13.3% 15.2% 16.9% 24.6% 16.7% 31.4% 20.8% 30.5% 30.5% 9.2% Interval) OVERALL TOTAL ACCURACY = 50.3% OVERALL KAPPA INDEX = 47.5% OVERALL 90% UPPER AND LOWER CONFIDENCE INTERVAL: 47.9% and 52.7% Producers Accuracy (Omission Error) Confidence Interval is 90% two-sided limit

Table 16. Contingency table for fuzzy accuracy assessment level 5.

108 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park Commission Error Commission Fuzzy 4 +/ -(90%Conf. %Correct %Correct Interval) Interval)

Reference (Accuracy Assessment Field Data) Sum

1 2 4 5 6 7 13 14 15 18 20 22 23 24 26 32 33 34 35 36 38 41 43 120 121 141 142 161 162 163 164 190 191 400

1 38 6 0 0 0 9 2 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 59 64.4% 11.1% 2 1 14 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 20 70.0% 19.4% 4 0 0 35 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 37 94.6% 7.5%

5 0 0 1 15 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 18 83.3% 17.2% Users Accuracy (Commission Error) Conf 6 0 0 0 0 31 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 33 93.9% 8.3% 7 11 1 0 0 0 32 5 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 51 62.7% 12.1% 13 2 0 0 0 0 1 21 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 7 1 0 0 0 1 0 0 0 0 3 37 56.8% 14.7% 14 0 0 1 0 0 0 0 24 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 28 85.7% 12.7% 15 0 0 0 0 0 0 0 1 17 5 0 0 0 0 0 3 0 0 0 0 0 0 0 0 1 0 0 1 0 1 0 1 0 0 30 56.7% 16.5% 18 0 0 0 0 0 0 0 0 6 24 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 75.0% 14.2$ 20 0 0 0 0 0 1 0 0 0 0 39 2 3 2 2 0 0 1 1 5 0 0 0 0 0 0 0 1 0 0 0 0 0 0 57 68.4% 11.0% 22 0 0 0 0 0 0 0 0 0 0 0 59 4 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 1 67 88.1% 7.3% 23 0 0 0 0 0 0 0 0 0 0 0 4 21 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 25 84.0% 14.1% 24 0 0 0 0 0 0 0 0 0 0 1 0 0 38 0 0 0 0 0 2 1 0 0 0 0 0 0 0 1 0 0 0 0 0 43 88.4% 9.2% 26 0 0 0 0 0 0 0 0 0 1 2 4 0 0 20 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 28 71.4% 15.8% 32 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 25.0% 48.1% 33 1 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 24 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 30 80.0% 13.7% 34 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 28 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 31 90.3% 10.3% 35 0 0 0 0 0 0 0 0 0 0 4 0 1 1 0 0 0 0 25 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 31 80.6% 13.3% idence Interval is 36 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 100.0% 1.6% 38 0 0 0 0 0 0 0 0 0 0 3 4 1 1 1 0 0 0 0 0 23 0 0 0 0 1 0 0 0 0 0 0 0 0 34 67.6% 14.7%

Sample Data (Polygon Map Data) 41 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 3 1 2 1 47 0 0 0 0 0 0 0 0 1 0 0 0 57 82.5% 9.2% 43 0 0 0 0 0 0 0 0 0 0 2 6 0 1 0 0 0 0 0 1 0 0 9 0 0 0 0 0 0 0 1 1 1 0 22 40.9% 19.5% 120 2 0 0 0 0 6 2 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 29 0 0 0 0 0 0 0 0 0 1 41 70.7% 12.9% 121 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 28 0 0 0 1 1 0 0 0 0 32 87.5% 11.2% 90% two-sided limit two-sided 90% 141 0 0 1 0 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 27 0 0 0 0 0 0 0 0 33 81.8% 12.6% 142 0 0 0 0 0 0 0 5 0 0 1 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 21 0 0 0 0 0 0 0 29 72.4% 15.4% 161 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 1 0 0 4 0 0 0 0 0 0 0 17 1 1 0 0 0 0 27 63.0% 17.1% 162 0 0 0 0 0 0 0 0 0 0 1 2 1 0 1 0 0 0 0 0 0 0 1 0 0 0 0 1 25 1 1 0 0 0 34 73.5% 13.9% 163 0 0 0 0 0 0 0 0 0 0 0 3 0 2 1 0 0 0 0 1 0 0 0 0 0 0 0 0 2 14 1 1 0 0 25 56.0% 18.3% 164 0 0 0 0 0 0 0 0 0 0 1 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 24 0 0 0 30 80.0% 13.7% 190 0 0 0 0 0 0 0 0 0 0 0 17 2 3 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 20 0 1 44 45.5% 13.5% 191 0 0 0 0 0 0 0 0 0 0 3 1 0 0 1 0 0 2 0 6 0 0 1 0 0 0 0 0 1 0 0 0 8 0 23 34.8% 18.5% 400 0 0 0 0 0 1 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 55 61 94.2% 7.1%

Sum 55 21 38 15 32 53 31 35 25 37 64 109 35 52 26 4 26 34 28 61 29 48 11 38 32 28 21 20 33 20 28 24 9 63 % Accurate 69.1% 66.7% 92.1% 100.0% 96.9% 60.4% 67.7% 68.6% 68.0% 64.9% 60.9% 54.1% 60.0% 73.1% 76.9% 25.0% 92.3% 82.4% 89.3% 52.5% 79.3% 97.9% 81.8% 76.3% 87.5% 96.4% 100.0% 85.0% 75.8% 70.0% 85.7% 83.3% 88.9% 87.3% Total Correct = 885 +/ - (90% Total Samples = 1185 Conf. 11.2% 19.3% 8.5% 3.3% 6.6% 12.0% 15.4% 14.3% 17.3% 14.3% 10.8% 8.3% 15.1% 11.1% 15.5% 48.1% 10.5% 12.2% 11.4% 11.3% 14.1% 4.4% 23.7% 12.7% 11.2% 7.6% 2.4% 15.6% 13.8% 19.4% 12.7% 14.6% 22.8% 7.7% Interval) OVERALL TOTAL ACCURACY = 74.7% OVERALL KAPPA INDEX = 72.4% OVERALL 90% UPPER AND LOWER CONFIDENCE INTERVAL: 72.6% and 76.8% Producers Accuracy (Omission Error) Confidence Interval is 90% two-sided limit

Table 17. Contingency table for fuzzy accuracy assessment level 4.

109 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park Commission Error Commission Fuzzy 3 +/ -(90%Conf. %Correct %Correct Interval) Interval)

Reference (Accuracy Assessment Field Data) Sum

1 2 4 5 6 7 13 14 15 18 20 22 23 24 26 32 33 34 35 36 38 41 43 120 121 141 142 161 162 163 164 190 191 400

1 49 3 0 0 0 4 1 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 60 81.7% 9.1% 2 0 20 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 23 87.0% 13.7% 4 0 0 35 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 37 94.6% 7.5%

5 0 0 1 15 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 18 83.3% 17.2% Users Accuracy (Commission Error) Conf 6 0 0 0 0 33 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 33 100.0% 1.5% 7 10 1 0 0 0 36 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 51 70.6% 11.5% 13 2 0 0 0 0 1 30 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 2 37 81.1% 11.9% 14 0 0 1 0 0 0 0 26 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 29 89.7% 11.0% 15 0 0 0 0 0 0 0 1 28 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 30 93.3% 9.2% 18 0 0 0 0 0 0 0 0 0 30 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 93.8% 8.6% 20 0 0 0 0 0 0 0 0 0 0 48 0 3 1 2 0 0 1 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 57 84.2% 8.8% 22 0 0 0 0 0 0 0 0 0 0 0 66 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 67 98.5% 3.2% 23 0 0 0 0 0 0 0 0 0 0 0 3 22 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 25 88.0% 12.7% 24 0 0 0 0 0 0 0 0 0 0 1 0 0 41 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 43 95.3% 6.4% 26 0 0 0 0 0 0 0 0 0 1 0 4 0 0 22 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 28 78.6% 14.5% 32 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 25.0% 48.1% 33 1 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 26 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 30 86.7% 11.9% 34 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 30 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 31 96.8% 6.8% 35 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 29 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 31 93.5% 8.9% idence Interval is 36 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 100.0% 1.6% 38 0 0 0 0 0 0 0 0 0 0 1 3 0 1 0 0 0 0 0 0 28 0 0 0 0 1 0 0 0 0 0 0 0 0 34 82.4% 12.2%

Sample Data (Polygon Map Data) 41 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 3 1 2 1 47 0 0 0 0 0 0 0 0 1 0 0 0 57 82.5% 9.2% 43 0 0 0 0 0 0 0 0 0 0 2 6 0 1 0 0 0 0 0 1 0 0 9 0 0 0 0 0 0 0 1 1 1 0 22 40.9% 19.5% 120 2 0 0 0 0 6 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 31 0 0 0 0 0 0 0 0 0 1 41 75.6% 12.3%

121 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 30 0 0 0 1 0 0 0 0 0 32 93.8% 8.6% 90% two-sided limit two-sided 90% 141 0 0 1 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 29 0 0 0 0 0 0 0 0 33 87.9% 10.9% 142 0 0 0 0 0 0 0 5 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 22 0 0 0 0 0 0 0 29 75.9% 14.8% 161 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 24 0 0 0 0 0 0 27 88.9% 11.8% 162 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 33 0 0 0 0 0 34 97.1% 6.2% 163 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 23 0 0 0 0 25 92.0% 10.9% 164 0 0 0 0 0 0 0 0 0 0 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 27 0 0 0 30 90.0% 10.7% 190 0 0 0 0 0 0 0 0 0 0 0 4 2 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 34 0 1 44 77.3% 11.5% 191 0 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 15 0 23 65.2% 18.5% 400 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 61 61 100.0% 0.8%

Sum 64 24 38 15 33 49 35 35 29 36 62 91 31 52 24 1 27 34 31 51 29 48 9 31 31 30 22 24 35 23 29 35 16 66 % Accurate 76.6% 83.3% 92.1% 100.0% 100.0% 73.5% 85.7% 74.3% 96.6% 83.3% 77.4% 72.5% 71.0% 78.8% 91.7% 100.0% 96.3% 88.2% 93.5% 62.7% 96.6% 97.9% 100.0% 100.0% 96.8% 96.7% 100.0% 100.0% 94.3% 100.0% 93.1% 97.1% 93.8% 92.4% Total Correct = 885 +/ - (90% Total Samples = 1185 Conf. 9.5% 14.6% 8.5% 3.3% 1.5% 11.4% 11.2% 13.6% 7.3% 11.6% 9.5% 8.2% 15.0% 10.3% 11.4% 50.0% 7.8% 10.6% 8.9% 12.1% 7.3% 4.4% 5.6% 1.6% 6.8% 7.1% 2.3% 2.1% 7.9% 2.2% 9.5% 6.1% 13.1% 6.1% Interval) OVERALL TOTAL ACCURACY = 86.7% OVERALL KAPPA INDEX = 80.9% OVERALL 90% UPPER AND LOWER CONFIDENCE INTERVAL: 85% and 88.4% Producers Accuracy (Omission Error) Confidence Interval is 90% two-sided limit

Table 18. Contingency table for fuzzy accuracy assessment level 3.

110 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 19. Map unit accuracies for omission and commission errors for all errors at all fuzzy levels (only vegetated polygons).

Map Unit Name Fuzzy 5 Fuzzy 4 Fuzzy 3 VEG CODE Users' accuracy Producers' accuracy Users' accuracy Producers' accuracy Users' accuracy Producers' accuracy 1 Herbaceous Upland Alpine > 9600 ft 52.5% 60.8% 64.4% 69.1% 81.7% 76.6% 2 Herbaceous Upland Alpine Fellfield 41.2% 41.2% 70.0% 66.7% 87.0% 83.3% 4 Herbaceous Upland Montane < 9600 ft 75.7% 71.8% 94.6% 92.1% 94.6% 92.1% 5 Herbaceous Wetland Cross Zone - Marsh 38.9% 100.0% 83.3% 100.0% 83.3% 100.0% 6 Herbaceous Wetland Cross Zone - Wetland 72.7% 54.5% 93.9% 96.9% 100.0% 100.0% 7 Herbaceous Wetland Subalpine / Alpine - Alpine Mead 45.1% 44.2% 62.7% 60.4% 70.6% 73.5% 13 Shrub Upland Alpine 35.1% 43.3% 56.8% 67.7% 81.1% 85.7% 14 Shrub Upland Lower Montane - Undifferentiated 14.3% 17.4% 85.7% 68.6% 89.7% 74.3% 15 Riparian Aspen 40.0% 54.5% 56.7% 68.0% 93.3% 96.6% 18 Upper Montane Aspen 62.5% 50.0% 75.0% 64.9% 93.8% 83.3% 20 Lower Montane Douglas-fir 26.3% 30.0% 68.4% 60.9% 84.2% 77.4% 22 Subalpine Mixed Conifer 88.1% 39.6% 88.1% 54.1% 98.5% 72.5% 23 Lodgepole Pine - High Elevation > 9500 ft 68.0% 43.6% 84.0% 60.0% 88.0% 71.0% 24 Lodgepole Pine - Low Elevation < 9500 ft 79.1% 54.0% 88.4% 73.1% 95.3% 78.8% 26 Lodgepole Pine - Rock 32.1% 56.3% 71.4% 76.9% 78.6% 91.7% 32 Cottonwood 25.0% 25.0% 25.0% 25.0% 25.0% 100.0% 33 Juniper 43.3% 81.3% 80.0% 92.3% 86.7% 96.3% 34 Ponderosa pine Graminoid 45.2% 32.6% 90.3% 82.4% 96.8% 88.2% 35 Ponderosa pine Rockland 32.3% 50.0% 80.6% 89.3% 93.5% 93.5% 36 Ponderosa pine Shrubland 59.4% 19.4% 100.0% 52.5% 100.0% 62.7% 38 Subalpine Limber Pine 35.3% 54.5% 67.6% 79.3% 82.4% 96.6% 41 Disturbance - Dead and Down 82.5% 97.9% 82.5% 97.9% 82.5% 97.9% 43 Blue Spruce 31.8% 58.3% 40.9% 81.8% 40.9% 100.0% 120 Shrub Riparian Cross Zone > 9600 ft 53.7% 55.0% 70.7% 76.3% 75.6% 100.0% 121 Shrub Riparian Cross Zone < 9600 ft 84.4% 75.0% 87.5% 87.5% 93.8% 96.8% 141 Shrub Upland Lower Montane - Big Sagebrush 57.6% 82.6% 81.8% 96.4% 87.9% 96.7% 142 Shrub Upland Lower Montane - Bitterbrush 44.8% 44.8% 72.4% 100.0% 75.9% 100.0% 161 Mixed Conifer with Aspen (Ponderosa pine) 18.5% 35.7% 63.0% 85.0% 88.9% 100.0% 162 Mixed Conifer with Aspen (Lodgepole Pine) 44.1% 50.0% 73.5% 75.8% 97.1% 94.3% 163 Mixed Conifer with Aspen (Douglas-fir) 12.0% 33.3% 56.0% 70.0% 92.0% 100.0% 164 Mixed Conifer with Aspen (Spruce - Fir) 23.3% 36.8% 80.0% 85.7% 90.0% 93.1% 190 Riparian Upper Montane Mixed Conifer > 8500 ft 13.6% 60.0% 45.5% 83.3% 77.3% 97.1% 191 Riparian Lower Montane Mixed Conifer < 8500 ft 17.4% 40.0% 34.8% 88.9% 65.2% 93.8% 400 Krummholz 77.0% 82.5% 90.2% 87.3% 100.0% 92.4% Mean 46.26% 52.23% 72.52% 76.94% 84.45% 89.89%

111 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Table 20. Change in mean map unit accuracies by fuzzy level.

Omission (Producers error) Commission (Users error) Fuzzy 5 52.2 46.3 Fuzzy 4 77 72.5 Fuzzy 3 90 84.4

112 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Accuracy Assessment by Map Unit.

Herbaceous Upland Alpine Both errors of omission and commission Map Unit – 1 are roughly the same.

Most of the errors for both omission and Herbaceous Upland Montane > 9600 commission in fuzzy class 5 are with ft. map unit 7 – Alpine Meadows. This Map Unit – 4 may be an artifact of the extremely dry year when the photographs were Most of the errors for both omission and acquired. Secondarily, errors for both commission in fuzzy class 5 are with the omission and commission are with map upland shrub map units 14, 141 and 142. unit 2 – Alpine Fellfield. Often, the The error appears to be a matter of shrub rocks one sees in fell-fields are small density, that is, low density shrub areas and thus fall below the photographic are often photo-interpreted as grain making them look like drier upland herbaceous. Fuzzy class 4 accuracy for herbaceous areas. Fuzzy class 4 this map unit improves significantly, as accuracy for this map unit improves AA points described as shrub are moved slightly as AA points described fellfield into this map unit. There is no are moved into this map unit. This trend additional increase in accuracy as on continues for fuzzy class 3. Both errors looks at fuzzy class 3. Both errors of of omission and commission are roughly omission and commission are roughly the same. the same.

Herbaceous Upland Alpine Fellfield Herbaceous Wetland Cross Zone - Map Unit – 2 Marsh Map Unit – 5 Most of the errors for both omission and commission in fuzzy class 5 are with There are no errors in this map unit. map unit 1 – Herbaceous Upland Alpine. This map unit is rare within the mapping Both omission and commission errors area but it does have a distinct signature are significant yielding only 41.2% for and is difficult to misinterpret. Some both. Often, the rocks one sees in AA points described wetlands or wet fellfields are small and thus fall below meadows, however, when one looks at the photographic grain. This then makes the imagery one finds the area sampled the signature much like drier upland to be underwater or shoreline. herbaceous areas. In addition, the Adjustments were made accordingly and photographic signature provided by field the photo-interpretation was assumed to plots shows a high amount of rock thus be absolutely correct – fuzzy class 5. confusing the interpretation. Fuzzy class 4 accuracy for this map unit improves slightly as AA points described as alpine meadow are moved into this map unit. This trend continues for fuzzy class 3.

113 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Herbaceous Wetland Cross Zone - respectively). These both improve Wetland markedly for fuzzy class 4 and Map Unit – 6 minimally more for fuzzy class 3 leaving this map unit below the mapping The errors in this map unit are almost accuracy norm of 80%. The greatest exclusively with other “wet” map units error for this map unit even after the such as alpine wet meadows – map unit application of fuzzy assessment is with 7, marshes – map unit 5 and riparian herbaceous upland meadows – map unit cross-zone shrublands – map unit 120. 1. The error associated with map units 5 and 7 are “degrees of wetness” Shrub Upland Alpine interpretations in the AA plots. In Map Unit – 13 addition, the photography was acquired during a drier than normal year and the The omission and commission errors in AA points were acquired during a wetter this map unit include alpine meadows than normal year creating a notable point and riparian shrublands. The errors of reference difference. The error appear to be both floristic and associated with the riparian shrubland physiognomic and are both quite high map unit is a density issue. Low-density (43.3% and 35.1% correct for errors of riparian shrublands are often interpreted omission and commission respectively). as wetlands. Error of omission is very The primary omission error is with wet high (44.2% correct) and reasonable meadows – map unit 7. This appears to (72.7% correct) for errors of commission be a density issue as described above in fuzzy class 5. These improve with other map units. Mapped (photo- markedly for fuzzy class 4 and are 100% interpreted) low density upland alpine correct for fuzzy class 3. shrublands mapped as herbaceous, in this case wet meadow, are often assigned Herbaceous Wetland Subalpine / to an upland shrub community. In Alpine – Alpine Meadow addition, many of the shrubs described Map Unit – 7 are either small in stature making them hard to discern on photographs or are The errors in this map unit are primarily heavily browsed. These errors improve with Herbaceous Alpine – map unit 1 somewhat as one moves from fuzzy and wetlands – map unit 6. This is not class 5 to fuzzy class 3. The other too surprising as this map unit sits source of error is confusion with riparian between the two in a wetness scale. shrub communities. Much of the error Secondarily, there are omission and differentiating riparian shrub commission errors with riparian communities from upland shrub shrublands – map unit 120. The error communities is the mental model used associated with the riparian shrubland by the photo-interpreters. The map unit is a density issue. Low-density assumption used was that riparian shrub riparian shrublands are often interpreted communities are usually found adjacent as wetlands or in this case, wet to water and this it seems is wrong. meadows. At fuzzy class 5 we have a Water percolating to the surface from low accuracy for both omission and springs or other sources quite distant commission (44.2% and 45.1% from water bodies will allow for the

114 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

development of riparian shrub improve to 65% and 75% for omission communities. At fuzzy class 3 this map and commission errors respectively at unit meets program standards. fuzzy level 4. With increasing relaxation of membership requirements, Shrub Upland Montane – we have 83% and 94% accuracies at Undifferentiated fuzzy level 3. Map Unit – 14 Montane Aspen Errors for this map unit are very high at Map Unit – 18 fuzzy class 5. Almost all omission and commission errors for this map unit are The greatest confusion here is with the with the two other shrub classes, Big other Aspen class (MU 15). At fuzzy Sagebrush – map unit 141 and level 5 we have 50% and 63% Bitterbrush – map unit 142. It appears accuracies for omission and commission that errors of omission are the greatest errors respectively. Errors of for this map unit with the both shrub commission also included confusion types and error of commission only with with the mixed conifer – aspen classes Bitterbrush. These errors improve (MU’s 161 and 163). Accuracies markedly and meet program norms as improve to 68% and 57% for omission one moves to fuzzy class 3. This is and commission errors respectively at because many shrub communities share fuzzy level 4. With increasing elements of both map unit 141 and 142. relaxation of membership requirements, The secondary error is that of we have 97% and 93% accuracies at commission with the Ponderosa pine fuzzy level 3. Shrub map unit – 36. This is a density issue as all of these errors occur in what Douglas-fir that AA crews called Ponderosa pine Map Unit – 20 woodland with very low tree density while the photo-interpreters called it Errors for this map unit at fuzzy level 5 shrubland and ignored the few trees in are very high and are spread across all the polygon. This map unit does not conifer types. The errors of omission meet program standards for error of and commission are almost equal, 30% omission but does for error of and 26.3% accurate respectively. The commission. accuracy improves markedly at fuzzy level 4. Both accuracies more than Riparian Aspen double. The error trend is the same as Map Unit – 15 errors are distributed across most conifer types. At fuzzy level 3 the accuracy is The greatest confusion here is with the acceptable for errors of commission and other Aspen class (MU 18). At fuzzy approaches minimum levels for omission level 5 we have 55% and 40% errors. The error in this map unit may be accuracies for omission and commission attributed to the great mixing found in errors respectively. Errors of Douglas-fir types in this area. All commission also included confusion vegetation plots and AA plots showed a with the mixed conifer – aspen classes high cover of other conifers leading to (MU’s 161 and 164). Accuracies

115 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park the confusion. In addition, most of the riparian influence extended. It now Douglas-fir occurs outside the Park appears that the riparian influence was boundary to the east. These areas were greatly overestimated. Commission not sampled during the initial vegetation error (high accuracy) is very low at all survey and therefore had few fuzzy levels. photointerpretive reference points. A high proportion of the errors in both Lodgepole Pine – High Elevation > omission and commission occurred with 9500 ft the Ponderosa pine map units. This is in Map Unit – 23 large part due to the recent encroachment of Douglas-fir into the This map unit shows low accuracy at Ponderosa pine zone. Often there fuzzy level 5 for errors of omission existed areas of Ponderosa pine making (44%) and reasonable errors for errors of up the primary canopy but with a high commission (68%). Most of these errors cover of Douglas-fir as a T2 layer. were with like map units such as Although this could be discerned on the Lodgepole – Rock (MU 26) and Mixed photographs, no map unit existed for this Lodgepole with Aspen (MU 162) occurrence as well as a recognized although there are other scattered errors vegetation description. for omission errors. Commission errors occurred almost exclusively with mixed Subalpine Mixed conifer lodgepole – spruce/fir stands. Class Map Unit – 22 accuracy improves in both fuzzy class 4 and 3 for omission error. Accuracy is This class shows a marked difference at consistently high for errors of fuzzy level 5 between omission and commission. No confusion is noted with commission errors. Errors of MU 24 as these are separated in the commission are minimal showing 88% model. accuracy while omission errors are high with accuracy of 40%. Most of the Lodgepole Pine – Low Elevation < omission error lies with misclassification 9500 ft of lodgepole pine, limber pine and the Map Unit – 24 riparian upper montane mixed conifer. The errors in the lodgepole pine area a Errors in this map unit mimic those in result of a mixed canopy of spruce, fir MU 23. Error typically occurs with and lodgepole. The error with these other like types however, it does occur classes drops notably in fuzzy level 4 with many other conifer types. and 3. Similar error was found with Omission error is fairly high and mixed canopy limber pine. The commission error is fairly low at fuzzy misclassification with the riparian mixed level 5. These both improve at fuzzy conifer type seems to be derived from an levels 4 and 3. However, omission error overestimation of the ecological extent does not quite meet map standards for of a riparian corridor. Many riparian errors of omission. polygons were delineated in drainages with a stream and the extent of the riparian influence estimated. There was no visible clue showing how far the

116 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Lodgepole Pine – Rock omission and commission (33% and Map Unit – 26 43% respectively). The accuracies both jump to map standard levels at fuzzy This map unit suffers low accuracy for level 4 (82% and 90% respectively) and both omission and commission at fuzzy improve slightly at fuzzy level 3 (88% level 5. The omission errors are evenly and 97% respectively). Most of the spread out amongst other conifer types. confusion at fuzzy level 5 was with like Commission errors occur mostly with types such as Ponderosa pine – Rockland lodgepole pine high elevation (MU 23) (MU 35), Ponderosa pine – Shrub (MU and the spruce/fir map unit (MU22). 36), Mixed Ponderosa pine – Aspen Accuracies improve markedly for fuzzy (MU 161) and the Douglas-fir type (MU levels 4 and 3. Accuracies for both 20) that had a mixed canopy with omission and commission are in the mid Ponderosa pine. 70% range at fuzzy level 4 and meet map accuracy standards for omission Ponderosa pine – Rockland errors at fuzzy level 3. Map Unit – 35

Cottonwood This map unit shows very low accuracy Map Unit – 32 at fuzzy level 5 for both errors of omission and commission (50% and The cottonwood map unit suffers from 32% respectively). The accuracies both very low occurrence and therefore jump to map standard levels at fuzzy estimations of map accuracy are suspect. level 4 (89% and 80% respectively) and However, these accuracies are low, most improve slightly at fuzzy level 3 (94% of the confusion occurring with riparian for both). Most of the confusion at Aspen (MU 15). fuzzy level 5 for errors of omission was with like types such as Ponderosa pine – Juniper Shrub (MU 36), Mixed Ponderosa pine – Map Unit – 33 Aspen (MU 161) and the Douglas-fir type (MU 20) that had a mixed canopy This map unit / association has very high with Ponderosa pine. Errors of accuracy (81%) for errors of omission commission were exclusively with other and low for errors of commission (43%) Ponderosa pine map units (MU 34 and at fuzzy level 5. The confusion at this 36) in addition to the Douglas-fir map level is generally with other Ponderosa unit (MU 20). pine types that have a mixed canopy of Ponderosa and Juniper. This is reflected Ponderosa pine - Shrubland by this map unit achieving map accuracy Map Unit – 36 standards at just fuzzy level 4. At fuzzy level 3 accuracies approach 100%. This map unit shows very low accuracy at fuzzy level 5 for both errors of Ponderosa pine - Graminoid omission and commission (19% and Map Unit – 34 54% respectively). The accuracies both jump at fuzzy level 4 to 53% and 100% This map unit shows very low accuracy respectively. Fuzzy level 3 shows a at fuzzy level 5 for both errors of moderate increase in the omission

117 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park accuracy to 63%. Most of the confusion Disturbance – dead and down at fuzzy level 5 for errors of omission Map Unit – 41 was spread amongst most conifer types. In most of these cases, there was a This map unit has high map standard mixed canopy with Ponderosa pine. accuracy at fuzzy level 5 and this does However, in this case we also see not change as one examines fuzzy levels confusion in the undifferentiated shrub 4 and 3. class (MU 14). This error is one of density differentiation. That is, the Blue Spruce observer on the ground providing the Map Unit – 43 reference data saw shrubland with a few scattered Ponderosas while the At fuzzy level 5 this map unit shows interpreter saw a low density Ponderosa very low accuracy. However, at fuzzy stand with shrubs. For this reason we level 5 this jumps considerably to 82% see the accuracy improve dramatically at accurate for omission error however fuzzy level 4 but still relatively low. commission error still remains low at Errors continue to occur with other map 41%. Error of omission is reduced units such as Douglas-fir (MU 20), further at fuzzy level 3 with a 100% Mixed Ponderosa pine – Aspen accuracy. Commission error remains (MU161) and some lodgepole. Errors of poor at 41%. Sample sizes are small for commission were exclusively with other this map unit. Ponderosa pine map units (MU 34 and 36) in addition to the Douglas-fir map Shrub Riparian – Cross Zone > 9600 unit (MU 20) at fuzzy level 4 and ft achieve 100% accuracy at fuzzy level 3. Map Unit – 120

Map Unit – 38 This map unit shows low accuracy at Subalpine Limber Pine fuzzy level 5 with 55% and 54% accuracies for errors of omission and This map unit shows low accuracy for commission respectively. At this level both errors of omission and commission the errors of omission are typically with (55% and 35% respectively) at fuzzy alpine shrublands (MU 13) and wetlands level 5. At this level, errors of omission (MU 6). Errors of commission include were spread out with no real marked alpine shrublands (MU 13), wetlands confusion with another class. The error (MU 6) and alpine wet meadows (MU of commission shows that most 7). At fuzzy level 4 the accuracies confusion occurred with the spruce-fir improve to 76% and 71% for omission map unit (MU 22) followed by Douglas- and commission respectively. The fir (MU 20) and lodgepole (MU’s 23 and omission error is primarily with alpine 24). Most of these confusions were shrublands (MU 13) while commission related to mixed canopies. At fuzzy error is primarily with alpine wet level 4 we see the accuracies jump to meadows (MU 7). The accuracies 79% and 68% for omission and improve to 100% and 76% for omission commission. At fuzzy level three, the and commission respectively at fuzzy accuracies reach 97% and 82% for level 3. omission and commission respectively.

118 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Shrub Riparian – Cross Zone < 9600 Shrub Upland Lower Montane – ft Bitterbrush Map Unit – 121 Map Unit – 142

Accuracies are reasonably high at fuzzy Accuracies are low for both errors of level 5 with 75% and 85% for omission omission and commission (45% for and commission respectively. The both) at fuzzy level 5. The omission confusion for omission error is typically confusion is primarily with with undifferentiated shrubland (MU undifferentiated shrublands (MU 14) and 14), mixed Ponderosa pine / Aspen (MU big sagebrush (MU 141). Commission 161) and riparian upper montane mixed confusion is primarily with conifer. Commission error is scattered undifferentiated shrublands (MU 14) but with no real trend. At fuzzy level 4 we some confusion also exists with have 88% accuracy for both errors of herbaceous montane uplands (MU 4) omission and commission. This and juniper woodlands (MU 33), both of improves to 97% and 94% accurate for which may have a shrub component. omission and commission respectively. This is reflected by the increase in accuracy at fuzzy level 4 to 100% and Shrub Upland Lower Montane – Big 72% for omission and commission Sagebrush respectively. At fuzzy level three Map Unit – 141 accuracies improve to 76% for errors of commission. Accuracies are reasonably high at fuzzy level 5 with 83% for omission errors. Mixed Conifer with Aspen (Ponderosa Commission errors are 58%. The pine) confusion for commission error is with Map Unit – 161 the Bitterbrush shrub map unit (MU 142). This is because the bitterbrush is Very low accuracies exist at fuzzy level often mixed with big sagebrush (MU 5 for this map unit (36% and 19% - 141). A large proportion of this error is omission and commission errors with the undifferentiated shrub map unit respectively). Omission errors generally (MU 14). Many of these errors occur on occurred with other conifer – aspen map the eastern portion of the mapping area units (MU’s 162, 163) but also other where there was no vegetation plot map aspen map units (MU’s 15 and 18). collection. In this area we find shrub Commission confusion is primarily with communities that were not originally the Ponderosa pine – graminoid map unit described such as Cercocarpus (MU 34). The confusion with MU 34 is montanus shrub types and these were primarily due to low density aspen in the often confused for bitterbrush. sampling area within a Ponderosa pine / Accuracies improve to 96.4% and 82% aspen polygon. This is reflected in fuzzy at fuzzy level 4 for omission and level 4 as we see accuracies improve commission respectively. Accuracies dramatically for errors of omission continue to improve to 97% and 88% at (85%) and 63% for commission errors. fuzzy level 3 for omission and Accuracy improves to 100% and 89% at commission respectively.

119 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park fuzzy level 3 for omission and commission errors. Accuracy improves commission errors respectively. to 100% and 92% at fuzzy level 3 for omission and commission errors Mixed Conifer with Aspen (Lodgepole respectively. Pine) Map Unit – 162 Mixed Conifer with Aspen (Spruce - fir) Very low accuracies exist at fuzzy level Map Unit – 164 5 for this map unit (50% and 44% - omission and commission errors Very low accuracies exist at fuzzy level respectively). Omission and 5 for this map unit (37% and 23% - commission errors are generally found omission and commission errors with other conifer – aspen map units respectively). Omission and (MU’s 161, 163,164) that have a commission errors are generally found considerable amount of mixed canopy. with other conifer – aspen map units Commission confusion is also found (MU’s 161, 162,163) that have a with the spruce – fir map unit (MU 22). considerable amount of mixed canopy. At fuzzy level 4 as we see accuracies Other omission errors are generally improve dramatically for errors of scattered with other map units. omission (76%) and 74% for Commission confusion is also found commission errors. Accuracy improves with the spruce – fir map unit (MU 22), to 94% and 97% at fuzzy level 3 for the low elevation lodgepole pine map omission and commission errors unit (MU 24) and the Douglas-fir map respectively. unit (MU 20). At fuzzy level 4 as we see accuracies improve for errors of Mixed Conifer with Aspen (Douglas- omission (86%) and for commission fir) (80%). Accuracy improves to 93% and Map Unit – 163 90% at fuzzy level 3 for omission and commission errors respectively. Very low accuracies exist at fuzzy level 5 for this map unit (33% and 12% - Riparian Upper Montane-Alpine omission and commission errors Mixed Conifer > 8500 ft. respectively). Omission and Map Unit – 190 commission errors are generally found with other conifer – aspen map units Accuracies are low for both errors of (MU’s 161, 162,164) that have a omission and commission (60% and considerable amount of mixed canopy. 14% for omission and commission Commission confusion is also found respectively) at fuzzy level 5. The with the spruce – fir map unit (MU 22) omission confusion is spread out and the low elevation lodgepole pine amongst several map units showing no map unit (MU 24). This low accuracy particular trend. Commission confusion reflects the generally poor ability to is occurs hugely with the spruce – fir differentiate Douglas-fir from other map unit (MU 22). This error is conifers. At fuzzy level 4 as we see discussed in more depth under the MU accuracies improve for errors of 22 description. Briefly, this error is due omission (70%) and 56% for

120 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park to unmet ecological justifications for Krummholz assuming the occurrence of this class. Map Unit – 400 Accuracies improve moderately at fuzzy level 4 yielding 83% and 46% accuracy The Krummholz map unit shows high for omission and commission errors accuracy at fuzzy level 5 with 83% and respectively. At fuzzy level three 77% accurate for errors of omission and accuracies improve to 97% and 77% for commission respectively. Most of the errors of omission and commission omission errors occurred with the alpine respectively. shrub map unit (MU 13). Commission errors occurred primarily with the spruce Riparian Lower Montane Mixed – fir map unit (22). The commission Conifer < 8500 ft. errors were due primarily to Map Unit – 190 interpretations of Krummholz height and the boundary between Krummholz and Accuracies are low for both errors of the spruce – fir forest. Accuracies omission and commission (40% and improve to 87% and 90% for fuzzy level 17% for omission and commission 4 and 92% and 100% for fuzzy level 3. respectively) at fuzzy level 5. Most of the error of omission occurs with Comparison of mean accuracies between confusion with the spruce- fir mixed fuzzy levels: Predictably, overall map canopy with aspen map unit (MU 164). accuracy increases as one relaxes Commission confusion is occurs with a requirements for individual map unit number of different map units. These membership. Table 15 shows the include the blue spruce map unit (MU increasing overall accuracy from fuzzy 43), Ponderosa pine – shrubland (MU level 5 to fuzzy level 3. 36) and the Douglas-fir map unit (MU 20). Accuracies improve markedly for Comparison of errors between fuzzy omission error at fuzzy level 4 yielding levels: As with overall map accuracy, 89%. Commission error remains results improve as per class requirements marginal at 35% accuracy. At fuzzy for membership within any particular level three accuracies improve to 94% class are relaxed. and 65% for errors of omission and commission respectively.

121 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Fuzzy Error Distribution

The distribution of error across the the minimum and maximum x and y for landscape may often provide additional each of the AA points and are based on valuable information. Certain zones may interpolations from data within the be more or less susceptible to error than mapping boundary. Low values indicate others. Portions of a vegetation map low accuracy. The values in the legend may be more accurate in some areas than reflect the fuzzy level designations. The others for any number of different kriging derived distribution map shows reasons. The error distribution is that accuracy is high throughout the completely unrelated to individual map entire area. There is no concentration of unit accuracies. The intent of the error error in any particular area. There is one distribution map is to provide additional “hole” in the north central portion of the insight to users. Figure 13 shows the mapping area in the , error distribution throughout the west of Fairchild Mountain and east of mapping area. Areas outside the the Desolation Peaks. These errors are mapping boundary may be ignored, as likely various map units being confused these are just calculations that extend to for bare rock.

122 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Figure 13. Fuzzy spatial view of accuracy.

123 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Discussion

NVC Classification the eastern buffer area and resolving some classification questions. In the As explained above, the vegetation of future it is recommended to have a ROMO vegetation mapping project is second field season after the draft very diverse ranging from the foothills to classification from the first field season alpine peaks, and although there has is completed. That way crews can target been previous local classifications and sampling of certain types to clarify vegetation research completed in and confusing types, augment under-sampled near the Park e.g., Marr 1977, types and look for un-sampled, but Komarkova 1979, Willard 1963 [alpine]; “expected” types. We also recommend Marr 1967, Peet 1981 forest and that preliminary map units (MU) be woodland]; Hess 1981, Hess and developed prior to the first field season Alexander 1986 [USFS Habitat Types]; and used to help allocate samples to MU Carsey et al. 2003, Cooper 1990, Kittel not well represented by NVC 1994 [wetlands], none were done in this associations. This will help insure comprehensive and detailed fashion for mappers have a minimum number of this large project/planning area. Even training site data to begin mapping. with all this previous work, several new NVC alliances and several new There are some unresolved classification associations were described and ranges issues that will need additional survey of several existing types expanded. work to further define such as clarifying some remaining sparse vegetation types Interestingly, there were a relatively high and classifying possible vegetation types number of NVC associations added to identified in the AA point data. the list of types in ROMO from the AA Anthropogenic disturbance of many of field work. Reasons for that include: 1) the lowland riparian vegetation types the 2002 vegetation crews focused their created challenges in classifying them. effort within the Park boundaries, but the AA was done for the whole project area, The plot data collected during this which included the 6 mile eastern buffer project was extensive with plots, area. In this lower elevation zone, field observation points and AA points crews encountered several foothill combining to total over 1800 new associations that likely do not occur sample sites with vegetation, within the Park. 2) Additionally, environmental and fuels data and photos. Krummholz and aspen-conifer mixed The data create a new “baseline” from forests types were not included in the which to evaluate past and future preliminary classification so 2002 field management issues and will be useful crews did not sample these types. These for years to come for various planning were later added as map units and and resource management activities consequently, showed up in the AA data. including fuels and fire management. NatureServe and CNHP ecologists did spend time in field during fall 2004 In the future, resource management augmenting the classification by personnel may key habitat for species of targeting sampling of the “new” types in concern to association, then locate

124 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park potential sites by using vegetation map The plot and AA data collected for the and environmental variables (e.g. North vegetation mapping project specifically slope Subalpine forest). identified occurrences of some tracked elements within the Park. Plots that are Global Rarity classified to a tracked element have been converted to an element occurrence National Parks such as ROMO play an using a generalized method to assign an important role in the global effort to element occurrence rank. These conserve biological diversity. It is a goal interpreted occurrences include both of ROMO management and planning to species identified on the plot species lists protect in perpetuity the natural world as well as community elements with in the Park and to serve as a world identified when the plot was classified to leader in wilderness protection, a specific plant association. management, and education (ROMO 2005). As an area where management Several globally and state rare strives to maintain the natural landscape communities and species were located and preserve natural ecological within the ROMO project area. These processes, ROMO preserves examples of include 19 different G1G2-G2G3 rare communities and species that communities and 3 G2G3 plant species. outside of Park boundaries are The rarest community elements threatened by a wide range of human identified from the plot data were several pressures that diminish biological G1G2 Ranked associations. These diversity. included Hesperostipa comata Colorado Herbaceous Vegetation In several ways, the vegetation mapping (G1G2), Muhlenbergia montana - program contributes to the ability of the Hesperostipa comata Herbaceous Parks to manage their landscapes for Vegetation (G1G2), Populus tremuloides conservation of biological diversity. The / Acer glabrum Forest (G1G2), and classification completed for the mapping Purshia tridentata / Artemisia frigida / projects identifies Park vegetation to the Hesperostipa comata Shrubland (G1G2). association level of the NVC. In total, 118 G1G2-G2G3 community Associations tracked by the Heritage element occurrences were identified Program are listed in the classification from the plot data. Table 21 presents all for the Park. Map units on the of the Natural Heritage association completed vegetation map likely to elements with an element rank of G1G2- contain the highest priority of these G2G3 identified from the ROMO plot elements can either be specifically data. surveyed to verify the presence of a rare community type, or these areas can be The rarest vascular plant species managed with consideration of the rare identified from the plot data was the G2 type in mind. Also, some occurrences of ranked Delphinium ramosum var. rare elements have been directly alpestre. Other important vascular plant identified from the plot data collected for elements identified in the plot data the classification and the accuracy included Castilleja puberula (G2G3) and assessment. Draba crassa (G3). In total, 37 G2-G3

125 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park vascular plant element occurrences were elements identified from the ROMO plot identified from the plot data. Table 21 data. lists the Natural Heritage plant

Table 21. Natural Heritage Elements Identified from Plot Data.

ELCODE SCIENTIFIC NAME G RANK S RANK PLOT CODE Natural Communities Hesperostipa comata Colorado Front Range Herbaceous CEGL001702 Vegetation G1G2 (blank) ROMOAA.194 ROMO.487, ROMO.804, ROMO.808, Muhlenbergia montana - ROMO.834, ROMO.839, Hesperostipa comata Herbaceous ROMOAA.176, ROMOAA.5783, CEGL001647 Vegetation G1G2 (blank) ROMOAA.5785, ROMOAA.719, ROMO.122, ROMO.485, ROMO.708, Populus tremuloides / Acer ROMOAA.1164, ROMOAA.24, CEGL000563 glabrum Forest G1G2 (blank) ROMOAA.5586, ROMOAA.961, Purshia tridentata / Artemisia frigida / Hesperostipa comata ROMO.8013, ROMOAA.1067, CEGL001055 Shrubland G1G2 (blank) ROMOAA.5168, Carex limosa Herbaceous CEGL001811 Vegetation G2 ROMO.824 Cercocarpus montanus / ROMOAA.558, ROMOAA.5652, CEGL001092 Hesperostipa comata Shrubland G2 (blank) ROMOAA.657 Juniperus scopulorum / CEGL000745 Cercocarpus montanus Woodland G2 (blank) ROMOAA.639 ROMO.402, ROMO.623, ROMO.628, ROMO.631, ROMO.828, ROMOAA.326, ROMOAA.328, ROMOAA.520, ROMOAA.5201, ROMOAA.5638, ROMOAA.5640, ROMOAA.5645, ROMOAA.5647, Juniperus scopulorum / Purshia ROMOAA.5651, ROMOAA.930, CEGL000749 tridentata Woodland G2 (blank) ROMOAA.936, ROMOAA.973, Pinus ponderosa / Alnus incana CEGL002638 Woodland G2 (blank) ROMOAA.738 ROMO.040, ROMO.108, ROMO.203, ROMO.515, ROMO.524, ROMO.620, ROMO.625, ROMO.829, ROMO.836, ROMO.840, ROMOAA.1097, ROMOAA.1113, ROMOAA.16, ROMOAA.173, ROMOAA.198, ROMOAA.267, ROMOAA.407, ROMOAA.415, ROMOAA.417, ROMOAA.418, ROMOAA.486, ROMOAA.5119, ROMOAA.5133, ROMOAA.5137, ROMOAA.519, ROMOAA.606, ROMOAA.729, ROMOAA.730, ROMOAA.801, Purshia tridentata / Muhlenbergia ROMOAA.827, ROMOAA.828, CEGL001057 montana Shrubland G2 (blank) ROMOAA.932 Rhus trilobata Rocky Mountain CEGL002910 Shrub Herbaceous Vegetation G2 (blank) ROMO.624 Ribes cereum / Leymus ambiguus CEGL001124 Shrubland G2 (blank) ROMOAA.760 CEGL003429 Glyceria grandis Herbaceous G2? (blank) ROMO.637, ROMOAA.79

126 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

ELCODE SCIENTIFIC NAME G RANK S RANK PLOT CODE Vegetation Picea engelmannii / Trifolium CEGL000377 dasyphyllum Forest G2? (blank) ROMO.199 Pseudotsuga menziesii / Carex CEGL000431 rossii Forest G2? (blank) ROMOAA.349, ROMOAA.5441, ROMO.046, ROMO.070, ROMO.173, ROMO.176, ROMO.278, ROMO.283, ROMO.491, ROMOAA.158, Danthonia intermedia Herbaceous ROMOAA.5835, ROMOAA.5926, CEGL001794 Vegetation G2G3 (blank) ROMOAA.869 Populus tremuloides - Pinus CEGL000540 flexilis Forest G2G3 (blank) ROMOAA.301, ROMOAA.994, ROMO.017, ROMO.031, ROMO.051, ROMO.059, ROMO.162, ROMO.271, ROMO.281, ROMO.286, ROMO.374, ROMO.445, ROMOAA.238, ROMOAA.245, ROMOAA.257, ROMOAA.260, ROMOAA.5033, ROMOAA.5034, ROMOAA.5050, Salix planifolia / Deschampsia ROMOAA.5057, ROMOAA.5108, CEGL001230 caespitosa Shrubland G2G3 (blank) ROMOAA.597, ROMOAA.604, Salix arctica - Salix nivalis ROMO.015, ROMO.295, ROMO.473, CEGL001432 Dwarf-shrubland G2Q (blank) ROMO.603, Vascular Plants PDRAN0B020 Delphinium ramosum var alpestre G2 S2 ROMO.107, ROMO.108 ROMO.238, ROMO.244, ROMO.248, ROMO.250, ROMO.258, ROMO.259, ROMO.384, ROMO.392, ROMO.423, ROMO.424, ROMO.426, ROMO.428, ROMO.429, ROMO.436, ROMO.437, ROMO.439, ROMO.441, ROMO.442, ROMO.443, ROMO.444, ROMO.449, ROMO.452, ROMO.453, ROMO.454, ROMO.455, ROMO.471, ROMO.473, ROMO.475, ROMO.476, ROMO.477, PDSCR0D2M0 Castilleja puberula G2G3 SNR ROMO.492, ROMO.496, ROMO.701 PDBRA110S0 Draba crassa G3 S3 ROMO.139, ROMO.337

Non-native Species accidentally as seed. Non-native species typically occur in and near areas where The vegetation of ROMO includes over human development has occurred, in 100 species that are considered non- areas that have been recently disturbed, native and that were either intentionally and along roads and trails. or un-intentionally introduced since before the time the Park was created in Non-native species have been 1915 (USDOI 2003). These non-native recognized as a serious threat to Park species may have been planted by early biodiversity and ecology. Of the 100 settlers in an attempt to improve pasture non native species known to occur in the forage, or may have arrived in the park Park, 35 have been targeted for control. To assist in the effort to control non-

127 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park native species in the Park, field crews 1.) Reduces the amount of time needed were instructed to note occurrences of for digital transfer or digitizing of the non-native species when observed, and line work; 2.) helps minimize shadows to record their presence when found in a and scale distortion in areas with large field plot. Although information changes in elevation; 3.) increases the regarding casual observance of an accuracy and thoroughness of the infestation was typically relayed to Park mapping by having recent, true-color managers it was not recorded and basemap imagery; 4.) allows for more formally transmitted or filed. useful and easier dissemination of draft Occurrences within plots were products to field crews, mappers, documented in the plot data and remain ecologists, etc., and 5.) is a great stand to be evaluated for possible management alone product that can be used in many action. other applications.

Aerial Photography and Photo-interpretation and Map Orthophotos Units

The acquisition of new orthophotos in Inherent to vegetation mapping projects addition to the aerial photography was is the need to produce both a consistent critical to our mapping efforts at ROMO. vegetation classification and a set of map We found that these not only saved time units. Typically, the systems are very in the digitizing and transfer stage but similar if not identical, but when using a also aided tremendously with map national classification such as the NVCS verification. The true color orthophotos there is typically not a strict one-to-one provided the utility of a map with the correspondence. This is due to the functionality of an aerial photo. In other remote sensing nature of photographic words, we could easily prepare and plot interpretation and its ability to only draft maps that contained both our delineate map units based on complex polygon outlines and a true color photo signatures. Subtle vegetation representation of the vegetation. In the characteristics that can be seen on the past, we would have had to either plot ground are not necessarily the same as polygons on less-clear black-and-white those apparent on the photos. Canopy orthophotos or use a clumsy closure, shadows, and timing of the combination of non-rectified aerial photography can also distort or obscure photos, simple color plots or digital photo signatures. raster graphics of the topographic maps of the area. Furthermore, as a digital For a highly diverse park such as ROMO product they afforded us the capability we suggest that a completed (or nearly of easily reproducing them for multiple completed) classification be in place users. before the actual interpretation begins. This is often difficult to do given the We would suggest that future projects constraints of time. Waiting for a strongly consider purchasing new complete classification of the vegetation orthophotos in addition to the aerial before proceeding with the mapping may photography for the following reasons: add a year or more to the entire process. The benefits include avoiding a revisit

128 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park or, in the worst case, redoing the reporting of accuracy data may lead to interpretation based on classification over or under estimation of map or map changes. Ideally, plot sampling should unit accuracy. Problems may arise from begin early in the project, followed by inaccurate reference data, data set miss- analysis of the vegetation data to the registration, poor or inappropriate NVC before the ground-truthing and sampling design, spatial variation of interpretation of the aerial photographs. accuracy, error magnitude and With this mapping effort a map unit procedural errors during the creation of meeting was held early in the project to the digital products. This project has try and determine appropriate map units. attempted to address these many pitfalls This was done using only “expected” and these problem areas are discussed associations and the input of Park staff below. and other ecologists. In addition, the lack of sampling to the east of the Park The term ground truth can be misleading boundary resulted in a reduced vision of as even classification of a location on the the associations that existed there in ground is subject to interpretation addition to a lack of photointerpretive (Foody 2001, Bird et al., 2002. The training sites. Another problem determination of vegetation association discerned late in the project was the using keys usually has some room for amount of mixing of conifer types. We interpretation of vegetative included no mixed conifer map units characteristics and even presence of other than the Subalpine Spruce – Fir species. The original vegetation map unit. Lodgepole mixed quite classification may have been developed extensively with all conifer types at high from samples collected during and low elevations. A mixed lodgepole significantly different climatic periods map unit would have been helpful. (e.g. wet year vs. dry year) or even seasonal variation (e.g., spring sampling Map Accuracy vs. fall sampling). A temporal change in the landscape between photo acquisition General Considerations / interpretation and field sampling for accuracy assessment is also common Judging the accuracy of a thematic map (Fire, landslides, avalanches etc.) has become as important as the actual Vegetation association descriptions also creation of that map yet the methods for depend heavily on estimations of cover collecting and interpreting accuracy that, in spite of extensive training prior assessment data remains problematic. to sampling, may be different enough to The concept of accuracy assessment is produce erroneous site classifications. straightforward however; the practicality For example, some AA plots were (measurement and expression) can be classified to the Spruce - Fir map unit tricky (Foody 2001). Foody (2001, based upon the cover value for spruce 2002) and even the Park mapping and fir. However, upon examination of protocols (Accuracy Assessment the field form we find that the cover type Procedures – 1994) discuss the many was actually co-dominant with limber sources of thematic error which may pine. If the map unit name assigned to lead to misinterpretations of accuracy the AA point was taken at face value assessments. The improper use or more than a few interpreted limber pine

129 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park polygons would have been wrong. This Given that source data may be rife with occurred on more than one occasion and problems, Foody (2001) suggests that emphasizes the need for a fuzzy the “…the typical accuracy assessment approach to map accuracy. is rather a measurement of the degree of agreement or correspondence between Exacerbating all of these potential the two data sets, rather than an problems is the underlying but false evaluation of the closeness of the assumption that the vegetation classes thematic map to reality.” This is are discrete rather than continuous. We probably the case with this project. know that rarely are vegetation types distinguished by sharp boundaries but Mis-registration of AA field samples and rather grade into one another (Gleason the actual location to be sampled can 1917, 1926, Whittaker 1956, 1962, cause tremendous problems and Curtis 1959). The degree of gradation invalidate an entire project. At ROMO, often will relate to the steepness of the we encountered a few points that were environmental gradient. “Steep miss-registered due to transcription environmental gradients tend to produce errors. One of the more common distinct vegetation boundaries where problems encountered was the location gradual environmental gradients tend to of plots within extremely small produce wider transition zones between polygons. Many polygons were well vegetation types.” (Standardized below the minimum mapping unit for the National Vegetation Classification project. Small polygons do provide for a System, 1994). Environmental gradients higher map precision however, the within ROMO vary from gradual to sampling of these can be very steep. Thus, the membership of a problematic. With very small polygons location or sample to a single discrete not only is the location a problem, vegetation type or description is suspect. especially in a forested site, but the edge The field key also assumes that any effect leads to considerable confusion in accuracy samples described in a plot classifying the area properly. have already been described when in Adjustments for this issue were reality a new association may be addressed during assessment of the confounding the classification in the fuzzy accuracy field data. field. Implicit is that the vegetation classification is complete and correct. Given the detail of the map and the Because the emphasis for this project is variability of the vegetation, we believe the vegetation map rather than the the accuracy assessment for ROMO was vegetation classification, no testing of successful due to several factors. First the classification has been conducted. In we made sure that the overall sampling a statistically perfect world, another design followed closely the protocols round of samples would have been described for by the National Park collected to test the vegetation Mapping Program. We did make some classification prior to any mapping. The allowances for practicality and statistical prohibitive costs for this test preclude it necessities. Boundaries between ever happening. polygons were minimally avoided but not so much so that only large homogenous areas were sampled. The

130 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park distribution of the sample points was introduced during the digitizing process. excellent and most map units received an These include: adequate number of points per type to 1. Lines are sometimes dropped draw general conclusions at each fuzzy between adjacent polygons and level. In addition the spatial distribution they may appear closed and of the AA sample points across the Park separate but in reality are not. was very good and the field crews made 2. The polygon coding may have every effort to reach many difficult sites. been incorrectly transcribed from Finally the predicted accuracy at almost the photo into the digital all un-sampled sites was high (Figure database. 13) and this tends to verify our 3. The polygon lines may have been assumption that no mapped area has a mis-registered to the base map. disproportionate amount of error. In other words, the transfer process from lines drawn on The overall magnitude of error for this mylar to a digital format requires project can be discerned by an a transformation from a non- examination of the contingency tables. rectified raster to a geo-rectified Typically, one will find errors between raster, followed by vectorization. similar map units. Occasionally one will The transformation process may find errors between such disparate introduce some mis-registration groups such as conifer map units and into the digital product. This shrub map units. At face value, this may problem is particularly noticeable be a gross error however; a closer look at with very small polygons or in the image and sample points often areas with extreme relief. clarifies the problem. For example, errors were often recorded when a crew The Accuracy Standard for the sampled a low density conifer type that NBS/NPS Vegetation Mapping may have been classified as either a Program shrub type with a scattering of conifers or a low density conifer plot with a shrub The program standards for accuracy are understory. When the AA crew reports 80% for both overall accuracy and their plot as a low-density conifer and individual class accuracy. The program the PI shows a shrub community, the recognizes that these levels of accuracy reporting of this type of error is may be difficult to achieve. Indeed, the maintained as wrong at fuzzy level 5 but Program Accuracy Assessment is accepted as correct at fuzzy level 4. Procedures states that “Given that This problem is typically one of vegetation mapping is necessarily perspective and is inherent in combining interpretive, it is recommended that ground information from remotely relaxed requirements be used in terms sensed information. of acceptable levels of error as well as confidence levels in the estimate. Very rarely did we find gross errors of Otherwise, regardless how carefully the classification without some explanation. mapping process is carried out, it is In these rare cases we speculate that the unlikely that accuracy requirements error is likely a result of inaccuracies will be met”. With the advent of fuzzy accuracy procedures we now have the

131 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park thematic map in several ways - very National Park (Cogan – pers comm.). stringent to relaxed. The choice of The overall fuzzy accuracies of these which standard to use would depend on parks are presented in Table 22 for the subject being mapped. In the case of comparative purposes. The trends across vegetation mapping, the preferred rigor this small set are similar. Given the would be relaxed vs. stringent. suggestions from the program standards and the results of four fuzzy accuracy When the vegetation mapping program assessments within the vegetation began, the use of fuzzy accuracy was mapping program we recommend that recognized but considered experimental the standard for stated and recognized with little use or publication. At this accuracy be fuzzy level 3. The point, eleven years after the publication definition for fuzzy level 3 as proposed of the Accuracy Assessment Procedures, by Gopal and Woodcock (1994) is much theory has been published but is “Reasonable or Acceptable Answer: remiss in applications. Even in the Maybe not the best possible answer but vegetation mapping program these it is acceptable; this answer does not techniques have not been used to any pose a problem to the user if it is seen on great extent. In addition to this project the map. Correct” therefore it would we know of three others that have used seem reasonable to accept this level as a fuzzy accuracy assessment (Hansen et al program standard. 2004, a, b, c) and informally at Zion

Table 22. Comparative fuzzy accuracies for four national parks.

Fuzzy 5 Fuzzy 4 Fuzzy 3 Rocky Mountain National Park 50.3% 74.7% 86.7% Walnut Canyon National Monument 50.0% 69.2% 96.9% Sunset Crater Volcano National Monument 53.9% 70.3% 86.8% Wupatki National Monument 59.1% 69.7% 92.2% Mean 53.3% 71.0% 90.7%

132 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Recommendations

and supplies. Many of the crew Field Survey members come from other areas of the country and often don’t yet Vegetation mapping is only as good as have a stable living situation. the fieldwork that backs it up. Since It’s hard for the crew’s to keep field crews are such an important their morale up if they have component of the project we found that nowhere to leave their making sure that they were well possessions, no where to shower, supported, well trained, and motivated and nowhere to call home. was instrumental to the success of the Housing needs to be available to project. The following are some them for the duration of the recommendations for ensuring that field summer. crews collect the best data possible. • Provide vehicles to the crews. Often the personal vehicles the • Start hiring early. By the end of crew members own are older and March most good field botanists less reliable than is a newer have already accepted a job. We government vehicle or a new started advertising for the rental vehicle. Very few positions in early December, and ecologists established in tried to complete the hiring by permanent positions would the end of February. consider using their own vehicle • Hire crew members for both on a daily basis for field work. botanical and outdoor skills. At There is no reason to expect ROMO crews had to backpack others to do differently. for many days at a time. If they • Provide the crews with the tools aren’t up for this, they have they need to do good work. difficulty collecting good data. Measuring tapes, plot markers, Make sure your crew members GPS units, cameras, dissecting are committed for the whole field scopes, field guides, and all season. incidental supplies (batteries, • Hire an even number of people pens, forms, files, folders) are all and have them work in crews of necessary and reasonable project two. A crew of two is typically costs. more efficient than two solo • Pay crews per diem for all field crews. This is also better from a days. safety standpoint since no one • Provide the crews with adequate has to work alone. training and orientation. Spend • Keep crew members for multiple the first week or two with them seasons. This reduces training in the field to ensure that all the time and adds consistency to field methods are understood and methods over the whole project. being followed. • Support your crews with • Meet regularly with your crews adequate housing, equipment, and make sure that they have everything that they need and are

133 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

following the methodology. be done with caution unless the results Encourage the crews to are verified by an associated accuracy document anything they assessment. General trends in vegetation encounter or are unsure of by (map unit) distribution may be expected taking extensive notes. These are but not hold up to rigorous testing. For very valuable in December when example, we attempted to model map you are trying to interpret what unit 22 – Spruce-Fir, with the they were seeing. assumption that north and south facing slopes should have distinct groups of Vegetation Map associations. Therefore, this map unit was divided along ridgelines during the The amount of mixed vegetation types at photointerpretation and assigned ROMO complicated the interpretation. separate map units with the assumption Lodgepole pine mixed freely at all that the accuracy assessment would elevations with all other conifer types. validate our mental model. This mental During the initial map unit designation model failed after examination of the meeting this issue did not come up and accuracy assessment results. The north therefore was not included as a map unit. and south facing Spruce-Fir polygons Had this been anticipated many mixed were then reassigned back to one map type polygons could probably been unit. Typically, one would then dissolve interpreted with some success. the lines between adjacent north and south facing Spruce-Fir polygons to Field sampling was restricted to those produce one polygon representing the areas within the Park. The result of this one map unit. In this case we was poor characterization of the maintained the separation of these vegetation in the large buffer area to the polygons for two reasons. One is that east of the Park. For example, there are the mental model is reasonable and that several more shrub types in the lower perhaps we just did not have enough elevations that were not described at all, information to properly model north and during the first sampling effort. In south facing Spruce-Fir map units. The addition, no sample points to the east of other reason is that maintaining the the Park also hurt the photointerpretive separation of polygons along the effort for lack of training sites. This was ridgelines may be useful for fire ameliorated somewhat by extensive modeling. visits to this area. Accuracy Map Units There are a number of areas that would benefit from revised protocol. ROMO map units contain varying levels Improvements in sampling and the data of detail. If more data becomes evaluation for fuzzy accuracy are two available about the distribution of certain areas that need the most work. The associations then additional modeling sample selection process allowed a should be attempted in the future. number of unanticipated problems to However, modeling associations and the creep in. Because of the excellent map units that encompass them should imagery, many polygons went well below the minimum mapping unit size.

134 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

The proportion of very small polygons Map Improvement Suggestions: increased which then gave them a disproportionate amount of weight 1. We would like to see the map during the sampling process. As periodically and formally refined discussed before, sampling these small and updated . This could be as polygons is problematic not only for simple as having field crews GPS locating and selecting the site properly record the locations of unique but also for edge effect of the vegetation not already on the surrounding vegetation. Large, more map or as involved as a new homogenous polygons should have photo interpretation effort. On received a greater proportion of the the low-cost side, the current sample points. This effectively provides vegetation map could help target an overly conservative estimation of likely stands within certain map overall and individual map unit classes and an efficient ground accuracy. truthing of just these types could follow. Through smaller scale The review process of each plot in order accuracy assessment and to provide a fuzzy designation was very verification efforts important time consuming and unanticipated. Each types such as rare and threatened plot had to be looked at and discussed communities and plant species amongst three people. This took a could be further defined. More tremendous amount of time. Some ideas costly efforts such as re-mapping on speeding this up include assigning an the entire Park are probably more additional secondary or alternate appropriate on a 10-25 year association to the point. This secondary timeframe. association, once assigned to its map 2. In addition to formal ground unit could then automatically receive a truthing we would also like to see fuzzy designation of four. This would more verification done by be similar to those AA points that piggybacking it onto other exactly matched the polygon designation projects. As opportunities arise, receiving a fuzzy designation of five. maps should be sent into the field This would greatly reduce the number of not only to be used but so they plots that had to be reviewed. Similar can be checked by competent problems were encountered when Zion crews. We encourage ROMO National Park went through the same and all researchers to continuely process (Cogan pers. com.). The fuzzy ground truth the map as they use review was necessarily subjective it. however it is conceivable that a numeric 3. All new vegetation data application can be devised that analyzes including GPS data and other the dataset and assign fuzzy levels based GIS layers should be wisely upon such items as cover, density, incorporated into this map. This wetland indication (FACW, FAC, OBL may involve such things as using etc) or other factor. new research that more accurately models certain vegetation types or updating the

135 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

current vegetation after a fire. The most straightforward method of Current advances in GIS and improving this map would be to GPS technology easily allows for incorporate the results of the accuracy updates to the digital map and assessment. This can be accomplished allows previous copies to be by recoding the inaccurate polygons to tracked and archived. Having an the appropriate map class as recorded on archive would allow for temporal the field form. Also general trends analyses such as examining observed on the contingency tables change over time and tracking could be included in the GIS layer. This the effects of climate change. may involve combing similar types or Overall, we feel strongly that this scaling the map classes up into broader product should not be static but categories change with new and better information.

136 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

BIBLIOGRAPHY

Accuracy Assessment Procedures, 1994. NBS/NPS Vegetation Mapping Program. United States Department of Interior, National Biological Survey and National Park Service

Anderson, H.E. 1982. Aids to determining fuel models for estimating fire behavior. USDA Forest Service, General Technical Report INT-GTR-122.

Alexander, R.R. 1985. Major Habitat Types, Community Types, and Plant Communities in the Rocky Mountains. USDA Forest Ser vice, General Technical Report RM-123.

Anderson, M., P. Bourgeron, M. T. Bryer, R. Crawford, L. Engelking, D. Faber- Langendoen, M. Gallyoun, K. Goodin, D. H. Grossman, S. Landaal, K. Metzler, K. D. Patterson, M. Pyne, M. Reid, L. Sneddon, and A. S. Weakley. 1998. International classification of ecological communities: terrestrial vegetation of the United States. Volume II. The National Vegetation Classification System: list of types. The Nature Conservancy, Arlington, Virginia, USA. 502 p.

Anderson, J.R., E.E. Hardy, J.T. Roach, R.E. Witmer. 1976. A land use and land cover classification system for use with remote sensor data. Geological Survey Professional Paper 964. U.S. Government Printing Office, Washington. 28pp.

Anderson, L. E. 1990. A checklist of Sphagnum In: North America north of Mexico. The Bryologist 93:500-501.

Anderson, L. E., H. A. Crum, and W. R. Buck. 1990. List of mosses of North America north of Mexico. The Bryologist 93:448-499.

Anderson M., P.S. Bourgeron, M.T. Bryer, R. Crawford, L. Engelking, D. Faber- Langendoen, M. Gallyoun, K. Goodin, D.H. Grossman, S. Landaal, K. Metzler, K.D. Patterson, M. Pyne, M. Reid, L. Sneddon, and A.S. Weakley. 1998. International Classification of Ecological Communities: Terrestrial vegetation of the United States. Volume II. The National Vegetation Classification System: List of types. The Nature Conservancy, Arlington, Virginia, USA.

Austin, M.P. and P.C. Heyligers. 1989. Vegetation survey design for conservation: gradsect sampling of forests in northeastern New South Wales, Biological Conservation. 50: 13-32.

137 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Bailey, R.G. 1995. Description of the ecoregions of the United States. 2d ed. Rev. and expanded (1st ed. 1980). Misc. Publ No. 1391 (rev.), Washington, DC: USDA Forest Service. 108 p. with separate map at 1:7,500,000.

Beidleman, Linda H., Biedleman, Richard G., & Willard, Beatrice E. (2000). Plants of Rocky Mountain National Park. Helena, MT: Falcon Publishing, Inc.

Billings W.D. 1988. Alpine Vegetation. Pages 391-420 in M. G. Barbour and W. D. Billings, editors. North American terrestrial vegetation. Cambridge University Press, New York.

Billings, W. D. and H. A. Mooney (1968). The ecology of arctic and alpine plants. Biological Review 43: 481-529.

Bird, A.C., Taylor, J.C. and Brewer, T.R. 2000. Mapping National Park landscape from ground, air and space. International Journal of Remote Sensing 21, 2719-36.

Braun-Blanquet, J. 1965. Plant sociology: the study of plant communities. (English translation of 2nd ed.) (Trans. rev. and ed. by C. D. Fuller and H. S. Conard) Hafner, London 439 p.

Brown, Peter M.; Kaufmann, Merrill R.; Shepperd, Wayne D. 1999. Long-term, landscape patterns of past fire events in a montane Ponderosa pine forest of central Colorado. Landscape Ecology. 14(6): 513-532.

Burgan, R.E. and R.C. Rothermel. 1984. BEHAVE: Fire behavior prediction and fuel modeling system – FUEL subsystem. USDA Forest Service, General Technical Report INT-GTR-167.

Carsey, K., G. Kittel, K. Decker, D.J. Cooper, and D. Culver. 2003. Field Guide to the Wetland and Riparian Plant Associations of Colorado. Colorado Natural Heritage Program, Colorado State University, Fort Collins, Colorado.

Chronic J. and H. Chronic. 1972. Prairie Peak and Plateau. Colorado Geological Survey Bulletin 32.

Clements, F.E. 1904. Formation and succession herbaria. University of Nebraska, University Studies IV (4):329-355.

Covington, W. W.; Moore, M. M. 1992. Postsettlement changes in natural fire regimes: implications for restoration of old-growth Ponderosa pine forests. In: Kaufmann, Merrill R.; Moir, W. H.; Bassett, Richard L., technical coordinators. Old-growth forests in the Southwest and Rocky Mountain regions: Proceedings of a workshop; 1992 March 9-13; Portal, AZ. Gen. Tech. Rep. RM-213. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station: 81-99.

138 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Curtis, J.T. 1959. The Vegetation of Wisconsin: An Ordination of Plant Communities. University of Wisconsin Press, Madison, Wisconsin.

Cogan, D, Reid, M, Schulz, K., Pucherelli, M. Zion National Park, Utah 1999-2003 Vegetation Mapping Project. 2003. Technical Memorandum 8260-03-01 Remote Sensing and GIS Group, Technical Service Center, Bureau of Reclamation, Denver, Colorado

Congalton, R. G. 1991. A review of assessing the accuracy of classifications of remotely sensed data. Remote Sensing of Environment 37:35-46.

Cowardin, L. M. V. Carter, F. C. Golet, and E. T. LaRoe. 1979. Classification of wetlands and deepwater habitats of the United States. U.S. Dep. Interior, Fish and Wildl. Serv. FWS/OBS - 79/31.

Daubenmire R.F. 1943. Vegetational Zonation in the Rocky Mountains. Bot. Rev. 9:325-393.

Daubenmire, R. 1952. Forest vegetation of northern Idaho and adjacent Washington, and its bearing on concepts of vegetation classification. Ecological Monographs 22:301 330.

Dickman, A. 1978. Reduced fire frequency changes species composition of a Ponderosa pine stand. Journal of Forestry 76: 24–25.

Driscoll, R. S., D. L. Merkel, D. L. Radloff, D. E. Snyder, and J. S. Hagihara. 1984. An ecological land classification framework for the United States. USDA Forest Service, Miscellaneous Publication No. 1439, Washington, D.C.

Egan, R. S. 1987. A fifth checklist of the lichen-forming, lichenicolous and allied fungi of the continental United States and Canada. The Bryologist 90:77-173.

Egan, R. S. 1989. Changes to the “Fifth checklist of the lichen-forming, lichenicolous and allied fungi of the continental United States and Canada”, edition I. The Bryologist 2:68-72.

Egan, R. S. 1990. Changes to the “Fifth checklist of the lichen-forming, lichenicolous and allied fungi of the continental United States and Canada”, edition II. The Bryologist 93:211-219.

Esslinger, T. L., and R. S. Egan. 1995. A sixth checklist of the lichen-forming, lichenicolous, and allied fungi of the continental United States and Canada. The Bryologist 98(4):467-549.

Eyre, F.H. (editor). 1980. Forest cover types of the United States and Canada. Society of American Foresters, Washington, D.C.

139 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Federal Geographic Data Committee. 1997. FGDC Vegetation Classification and Information Standards. Reston, VA.

Flahault, C. and C. Schroter. 1910. Rapport sur la nomenclature phytogeographique. Proceedings of the 3rd International Botanical Congress, Brussels, 1910, 1:131 164.

Fyfe, J.C. and G.M. Flato. 1999. Enhanced climate change and its detection over the Rocky Mountains. Journal of Climate, 12 (1), 230-243.

Foody, Giles M. 2001. GIS: the accuracy of spatial data revisited. In: Progress in Physical Geography, 25,3. 389-398.

Foody, G.M. (2002) Status of land cover classification accuracy assessment. Remote Sensing of Environment, 80 (1), 185-201.

Gillison, A.N. and K.R.W. Brewer. 1985. The use of gradient directed transects or gradsects in natural resource survey. Journal of Environmental Management. 20:103- 127.

Gleason, H.A. 1917. The structure and development of the plant association. Bulletin of the Torrey Botanical Club 44:463–481.

Gleason, H.A. 1926. The individualistic concept of the plant association. Bulletin of the Torrey Botanical Club 53:7–26.

Gopal, S., and C. Woodcock. 1994. Theory and methods for accuracy assessment of thematic maps using fuzzy sets. Photogrammetric Engineering & Remote Sensing 60, no. 2: 181-88.

Grossman, D.H., D. Faber-Langendoen, A.W. Weakley, M. Anderson, P. Bourgeron, R. Crawford, K. Goodin, S. Landaal, K. Metzler, K.D. Patterson, M. Pyne, M. Reid, and L. Sneddon. 1998. International Classification of Ecological Communities: Terrestrial Vegetation of the United States. Volume I: The National Vegetation Classification Standard. (Draft June 1997.) The Nature Conservancy, Arlington, VA.

Haase, R. 1987. An alpine vegetation map of Caribou Lake valley and Fourth of July Valley, Front Range, Colorado, U.S.A. Artic and Alpine Research, Vol. 19, No. 1, pp. 1 – 10.

Haeffner, Arden D. 1971. Daily temperatures and precipitation for Subalpine forests, Colorado. USDA Forest Service Research Paper RM-80, 48 p. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO.

Hansen M., Coles J., Thomas K.A., Cogan D., Reid M., Von Loh, J., Schulz K. 2004. USGS-NPS National Vegetation Mapping Program: Sunset Crater Volcano National Monument, Arizona, Vegetation Classification and Distribution. U.S. Geological Survey Southwest Biological Science Center.

140 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Hansen M., Coles J., Thomas K.A., Cogan D., Reid M., Von Loh, J., Schulz K. 2004. USGS-NPS National Vegetation Mapping Program: Wupatki National Monument, Arizona, Vegetation Classification and Distribution. U.S. Geological Survey Southwest Biological Science Center.

Hansen M., Coles J., Thomas K.A., Cogan D., Reid M., Von Loh, J., Schulz K. 2004. USGS-NPS National Vegetation Mapping Program: Walnut Canyon National Monument, Arizona, Vegetation Classification and Distribution. U.S. Geological Survey Southwest Biological Science Center.

Harris, A.G. and Tuttle, E.; 1990; Geology of the National Parks; Kendall/Hunt Publishers, Dubuque, IA.

Hauer, F.R., J.S. Baron, D.H. Campbell, K.D. Fausch, S.W. Hostetler, G.H. Leavesley, P.R. Leavitt, D.M. McKnight and J.A. Stanford. 1997. Assessment of climate change and freshwater ecosystems of the Rocky Mountains, USA and Canada. Hydrological Processes, 11, 903-924.

Hess, K. and R.A. Alexander. 1986. Forest Vegetation of the Arapaho and Roosevelt National Forests in Central Colorado: A Habitat Type Classification. USDA Forest Service Research Paper RM-266.

Hill, M. O. 1979. Twinspan, a Fortran program for arranging multivariate data in an ordered two-way table by classification of the individuals and attributes. Cornell Univ., Ithaca, NY. 90 p.

Hill, M. O., and H. G. Gauch, Jr. 1980. Detrended correspondence analysis: An improved ordination technique. Vegetatio. 42:47-58.

Hofmann, N., L. and others. 1998. Climate Change and Variability: Impacts on Canadian Water. Canada Country Study (CCS), Chapter 1, Vol. VIII, 1-120.

Hunter, M. 1991. Coping with ignorance: the coarse-filter strategy for maintaining biodiversity. Pages 256-281 in K.A. Kohm, ed., Balancing on the Brink of Extinction-the Endangered Species Act and Lessons for the Future. Island Press, Washington, DC.

Isard. S.A. 1986. Factors influencing soil moisture and plant community distributions on Niwot Ridge, Front Range, Colorado, U.S.A. Arct. Alp. Res. 18:83-96.

Jenkins, R. E. 1976. Maintenance of natural diversity: approach and recommendations. Pages 441-451 in Transactions of the 41st North American Wildlife Conference, 4 March 1976, Washington, DC. Wildlife Management Institute, Washington, DC.

Jenness, J. 2005. Surface Tools (surf_tools.avx) extension for ArcView 3.x, v. 1.6. Jenness Enterprises. Available at: http://www.jennessent.com/arcview/surface_tools.htm.

141 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Jenness, J. and J. J. Wynne. 2004. Kappa analysis (kappa_stats.avx) extension for ArcView 3.x. Jenness Enterprises. Available at: http://www.jennessent.com/arcview/kappa_stats.htm.

Jennings, M., O. Loucks, D. Glenn-Lewin, R. Peet, D. Faber-Langendoen, D. Grossman, A. Damman, M. Barbour, R. Pfister, M. Walker, S. Talbot, J. Walker, G. Hartshorn, G. Waggoner, M. Abrams, A. Hill, D. Roberts, and D. Tart. 2003. Guidelines for describing associations and alliances of the U.S. National Vegetation Classification. The Ecological Society of America, Vegetation Classification Panel, Version 3.0 November 2003. 100 pp. (+ Appendices)

Jensen, J. R. 1996. Introductory Digital Image Processing: A Remote Sensing Perspective (Second edition). Prentice Hall, Inc., Upper Saddle River, New Jersey, USA.

Kartesz, J.T. 1999. A synonymized checklist and atlas with biological attributes for the vascular flora of the United States, Canada, and Greenland. First edition. In: Kartesz, J.T. and C.A. Meacham. Synthesis of the North American flora [computer program]. Version 1.0. North Carolina Botanical Garden: Chapel Hill, NC.

Kaufmann, Merrill R., Paula J. Fornwalt, Laurie S. Huckaby, and Jason M. Stoker. 2001. Cheesman Lake - a historical pondersosa pine landscape guiding restoration in the South Platte Watershed of the Colorado Front Range. page 9 - 18 in Vance, RK; et al. Ponderosa pine ecosystems restoration an conservation proceedings RMRS-P-22. USDA Forest Service Rocky Mountain Research Station

Leith, R. M. M. and P. H. Whitfield. 1998. Evidence of climate change effects on the hydrology of streams in south-central B.C. Canadian Water Resources Journal, 23(3), 219-230.

McCabe, G.J. and D. M. Wolock. 1999. General-Circulation-Model Simulations of Future Snowpack in the , JAWRA, 35(6), 1473-1484.

Marr J.W. 1961. Ecosystems of the east slope of the Front Range of Colorado. University of Colorado Studies, Series in Biology, 8, 134 pp.

Marr, J. W. 1968a. Data on mountain environments II. Front Range, Colorado, four climax regions, 1953-1958. University of Colorado Studies Series in Biology No. 28, 171 p. University of Colorado Press, Boulder

Marr, J. W. 1968b. Data on mountain environments III. Front Range, Colorado, four climax regions, 1959-1964. University of Colorado Studies Series in Biology No. 29, 181 p. University of Colorado Press, Boulder

Maybury, K.P. 1999. Seeing the Forest and the Trees: Ecological classification for Conservation. The Nature Conservancy, Arlington, Virginia.

Mehl, M.S. 1992. Old-growth descriptions for the major forest cover types in the Rocky Mountain region. In: Kaufmann M.R., Moir W.H., Bassett R.L., Technical

142 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Coordinators. Old-growth forests in the southwest and Rocky Mountain regions. Proceedings of a workshop; Mar 9-1992 Mar 13; Portal, Arizona. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO: USDA Forest Service, General Technical Report RM-213. 106-120.

Merriam, C. H. and Steineger, L. 1890. Results of a biological survey of the San Francisco mountain region and the desert of the Little Colorado, Arizona. North American Fauna Report 3. U.S. Department of Agriculture, Division of Ornithology and Mammalia, Washington, D.C., 136 pp.

Moravec, J., 1993. Biodiversity changes on an ecosystemic level phytocoenological approach. Ekologia 12, 317–324.

Mueller-Dombois, D. and H. Ellenberg. 1974. Aims and methods of vegetation ecology. John Wiley and Sons, New York, New York. 547 pp.

NatureServe (2005, April 5) Classification of Standard Ecological Units (Online), 2005, May 20. http://www.natureserve.org/explorer/classeco.htm#vegetationClass

Neve, Lee A. 2000, Soil Survey of Rocky Mountain National Park, Colorado, Parts of Boulder, Grand, and Larimer Counties. United States Department of Agriculture, Natural Resources Conservation Service.

Odland, J. 1988. Spatial Autocorrelation. Newbury Park: SAGE.

Peet, R.K. 1978. Forest vegetation of the Colorado Front Range: Patterns of species diversity. Vegetatio 37: 65-78.

Peet, R.K. 1981. Forest vegetation of the Front Range: Composition and dynamics. Vegetatio 45: 3-75.

Peet, R. K. 1988. Forests of the Rocky Mountains. Pages 63-101 in M. G. Barbour and W. D. Billings, editors. North American terrestrial vegetation. Cambridge University Press, New York.

Pfister, R.D. and S.F. Arno. 1980. Classifying forest habitat types based on potential climax vegetation. Forest Science 26:52 70.

ROMO. 2005. Park Wilderness. Rocky Mountain National Park, National Park Service Web page. Viewed 7/25/2005 http://www.nps.gov/romo/resources/wilderness.html

Rondeau, Renée. 2001. Ecological System Viability Specifications for Southern Rocky Mountain Ecoregion. 1st edition. Colorado Natural Heritage Program.

Rydberg, P.A. 1914. Phytogeographical notes on the Rocky Mountain region III. Formations in the Alpine zone. Torrey Bot. Club, Bul. 41: 459-482.

143 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Rydberg, P.A. 1915. Phytogeographical notes on the Rocky Mountain region IV. Forests of the Subalpine and montane zones. Torrey Bot. Club, Bul. 42: 11-25.

Rydberg, P.A. 1915. Phytogeographical notes on the Rocky Mountain region V. Grasslands of the Subalpine and montane zones. Torrey Bot. Club, Bul. 42: 629-642.

Smith, W.K. 1985. Western montane forests, pp. 95-126 in B.F. Chabot and H.A. Mooney (eds.), Physiological ecology of North American plant communities. Chapman & Hall, New York.

Stohlgren, T. J., G. W. Chong, M. A. Kalkhan, and L. D. Schell. 1996. Rapid assessment of plant diversity patterns: A methodology for landscapes. Environmental Monitoring and Assessment 00: 1-19, 1996.

Stohlgren, T. J., G. W. Chong, M. A. Kalkhan, and L. D. Schell. 1997. Multiscale sampling of Plant Diversity: Effects of the minimum mapping unit. Ecological Applications 7(3): 1064-1074, 1997.

Stohlgren, T.J., A. Owen, and M. Lee. 2000. Monitoring shifts in plant diversity in response to climate change: a method for landscapes. Biodiversity and Conservation 9:65-86, 2000.

Stotler, R., and B. Crandall-Stotler. 1977. A checklist of liverworts and hornworts of North America. The Bryologist 80:405-428.

Taylor, A.M. 1999. Guide to the . Cataract Lode Mining Company. 221 pp.

The Nature Conservancy [TNC]. 1998. An environmentally-driven approach to vegetation sampling and mapping at Yosemite National Park. Report prepared for the U.S. Department of the Interior, National Biological Survey and National Park Service. The Nature Conservancy, Arlington, Virginia.

Thomas, K.A., and Jacobs, S.R. 2000. Analysis of Land Cover Map Accuracy in Northern Arizona. Final Report to the USGS National Gap Analysis Program. U.S. Dept. of the Interior.

Turner G. T. 1975. Mountain grassland ecosystem. USDA Forest Service Research Paper RM-161, Rocky Mt. For. and Range Esp. Stn., Fort Collins, CO.

USDA, NRCS. 2005. The PLANTS Database, Version 3.5 (http://plants.usda.gov). Data compiled from various sources by Mark W. Skinner. National Plant Data Center, Baton Rouge, LA 70874-4490 USA.

144 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

USDOI. 2003. Invasive Exotic Plant Management Plan and Environmental Assessment for Rocky Mountain National Park. United States Dept. of the Interior, National Park Service, Rocky Mountain National Park.

Veblen, T.T. 1998. Disturbance Patterns in Southern Rocky Mountain Forests. In R.L. Knight, F.W. Smith,S.W. Buskirk, W.H. Romme and W.L. Baker (editors), Forest Fragmentation in the . Island Press.

Westhoff, V. and E. van der Maarel. 1973. The Braun-Blanquet approach. Pp. 617 726 R.H. Whittaker (editor). Handbook of vegetation science, Part V: ordination and classification of communities. Junk, The Hague, The Netherlands.

Whittaker, R.H. 1956. Vegetation of the Great Smoky Mountains. Ecological Monographs 26:1 80.

Whittaker, R.H. 1962. Classification of natural communities. Botanical Review 28:1 239.

Williams, M.W., M. Losleben, N. Caine and D. Greenland. 1996. Changes in climate and hydrochemical responses in a high-elevation catchment in the Rocky Mountains, USA. Limnology and Oceanography, 41 (5), 939-946.

Zadeh, L. A. 1965. Fuzzy sets. Information and Control 8: 338-53.

145 USGS-NPS Vegetation Mapping Program Rocky Mountain National Park

Project DVD-ROM

146